Authors: Xin Hu, Yue Kang, Guanzi Yao, Tianze Kang, Mengjie Wang, Heyao Liu
Abstract: This study addresses the generalization limitations commonly observed in large language models under multi-task and cross-domain settings. Unlike prior methods such as SPoT, which depends on fixed prompt templates, our study introduces a unified multi-task learning framework with dynamic prompt scheduling mechanism. By introducing a prompt pool and a task-aware scheduling strategy, the method dynamically combines and aligns prompts for different tasks. This enhances the model's ability to capture semantic differences across tasks. During prompt fusion, the model uses task embeddings and a gating mechanism to finely control the prompt signals. This ensures alignment between prompt content and task-specific demands. At the same time, it builds flexible sharing pathways across tasks. In addition, the proposed optimization objective centers on joint multi-task learning. It incorporates an automatic learning strategy for scheduling weights, which effectively mitigates task interference and negative transfer. To evaluate the effectiveness of the method, a series of sensitivity experiments were conducted. These experiments examined the impact of prompt temperature parameters and task number variation. The results confirm the advantages of the proposed mechanism in maintaining model stability and enhancing transferability. Experimental findings show that the prompt scheduling method significantly improves performance on a range of language understanding and knowledge reasoning tasks. These results fully demonstrate its applicability and effectiveness in unified multi-task modeling and cross-domain adaptation.
Authors: Yue Zhang, Jiaxin Zhang, Qiuyu Ren, Tahsin Saffat, Xiaoxuan Liu, Zitong Yang, Banghua Zhu, Yi Ma
Abstract: We introduce \textbf{GAUSS} (\textbf{G}eneral \textbf{A}ssessment of \textbf{U}nderlying \textbf{S}tructured \textbf{S}kills in Mathematics), a benchmark that evaluates LLMs' mathematical abilities across twelve core skill dimensions, grouped into three domains: knowledge and understanding, problem solving and communication, and meta-skills and creativity. By categorizing problems according to cognitive skills and designing tasks that isolate specific abilities, GAUSS constructs comprehensive, fine-grained, and interpretable profiles of models' mathematical abilities. These profiles faithfully represent their underlying mathematical intelligence. To exemplify how to use the \textsc{GAUSS} benchmark, we have derived the skill profile of \textsc{GPT-5-thinking}, revealing its strengths and weaknesses as well as its differences relative to \textsc{o4-mini-high}, thereby underscoring the value of multidimensional, skill-based evaluation.
Authors: Haoyu Wang, Fengze Liu, Jiayao Zhang, Dan Roth, Kyle Richardson
Abstract: Event causality identification (ECI), a process that extracts causal relations between events from text, is crucial for distinguishing causation from correlation. Traditional approaches to ECI have primarily utilized linguistic patterns and multi-hop relational inference, risking false causality identification due to informal usage of causality and specious graphical inference. In this paper, we adopt the Rubin Causal Model to identify event causality: given two temporally ordered events, we see the first event as the treatment and the second one as the observed outcome. Determining their causality involves manipulating the treatment and estimating the resultant change in the likelihood of the outcome. Given that it is only possible to implement manipulation conceptually in the text domain, as a work-around, we try to find a twin for the protagonist from existing corpora. This twin should have identical life experiences with the protagonist before the treatment but undergoes an intervention of treatment. However, the practical difficulty of locating such a match limits its feasibility. Addressing this issue, we use the synthetic control method to generate such a twin' from relevant historical data, leveraging text embedding synthesis and inversion techniques. This approach allows us to identify causal relations more robustly than previous methods, including GPT-4, which is demonstrated on a causality benchmark, COPES-hard.
Authors: Seungyoun Yi, Minsoo Khang, Sungrae Park
Abstract: Automatic Prompt Optimization (APO) improves large language model (LLM) performance by refining prompts for specific tasks. However, prior APO methods typically focus only on user prompts, rely on unstructured feedback, and require large sample sizes and long iteration cycles-making them costly and brittle. We propose ZERA (Zero-init Instruction Evolving Refinement Agent), a novel framework that jointly optimizes both system and user prompts through principled, low-overhead refinement. ZERA scores prompts using eight generalizable criteria with automatically inferred weights, and revises prompts based on these structured critiques. This enables fast convergence to high-quality prompts using minimal examples and short iteration cycles. We evaluate ZERA across five LLMs and nine diverse datasets spanning reasoning, summarization, and code generation tasks. Experimental results demonstrate consistent improvements over strong baselines. Further ablation studies highlight the contribution of each component to more effective prompt construction. Our implementation including all prompts is publicly available at https://github.com/younatics/zera-agent.
Authors: Haodong Zhao, Chenyan Zhao, Yansi Li, Zhuosheng Zhang, Gongshen Liu
Abstract: The capacity of Large Language Models (LLMs) to reason is fundamental to their application in complex, knowledge-intensive domains. In real-world scenarios, LLMs are often augmented with external information that can be helpful, irrelevant, or even misleading. This paper investigates the causal impact of such auxiliary information on the reasoning process of LLMs with explicit step-by-step thinking capabilities. We introduce SciAux, a new dataset derived from ScienceQA, to systematically test the robustness of the model against these types of information. Our findings reveal a critical vulnerability: the model's deliberative "thinking mode" is a double-edged sword. While helpful context improves accuracy, misleading information causes a catastrophic drop in performance, which is amplified by the thinking process. Instead of conferring robustness, thinking reinforces the degree of error when provided with misinformation. This highlights that the challenge is not merely to make models "think", but to endow them with the critical faculty to evaluate the information upon which their reasoning is based. The SciAux dataset is available at https://huggingface.co/datasets/billhdzhao/SciAux.
Authors: Junlin Wang, Zehao Wu, Shaowei Lu, Yanlan Li, Xinghao Huang
Abstract: Retrieval-Augmented Generation (RAG) enables large language models (LLMs) to access external knowledge sources, but the effectiveness of RAG relies on the coordination between the retriever and the generator. Since these components are developed independently, their interaction is often suboptimal: the retriever may return irrelevant or redundant documents, while the generator may fail to fully leverage retrieved evidence. In this work, we propose a process-supervised multi-agent framework to bridge the gap between retriever and generator. The framework introduces two lightweight agents: a Decision Maker, which determines when to continue retrieval or stop for answer generation, and a Knowledge Selector, which filters retrieved documents to retain only the most useful evidence. To provide fine-grained supervision, we employ an LLM-as-a-Judge that evaluates each intermediate action with process-level rewards, ensuring more accurate credit assignment than relying solely on final answer correctness. We further adopt a tree-structured rollout strategy to explore diverse reasoning paths, and train both agents with Proximal Policy Optimization (PPO) in an end-to-end manner. Experiments on single-hop and multi-hop question answering benchmarks show that our approach achieves higher accuracy, more stable convergence, and produces more interpretable reasoning trajectories compared with standard RAG baselines. Importantly, the proposed framework is modular and plug-and-play, requiring no modification to the retriever or generator, making it practical for real-world RAG applications.
Authors: Aditi Debsharma, Bhushan Jagyasi, Surajit Sen, Priyanka Pandey, Devicharith Dovari, Yuvaraj V. C, Rosalin Parida, Gopali Contractor
Abstract: Emotion Recognition in Conversation has been seen to be widely applicable in call center analytics, opinion mining, finance, retail, healthcare, and other industries. In a call center scenario, the role of the call center agent is not just confined to receiving calls but to also provide good customer experience by pacifying the frustration or anger of the customers. This can be achieved by maintaining neutral and positive emotion from the agent. As in any conversation, the emotion of one speaker is usually dependent on the emotion of other speaker. Hence the positive emotion of an agent, accompanied with the right resolution will help in enhancing customer experience. This can change an unhappy customer to a happy one. Imparting the right resolution at right time becomes easier if the agent has the insight of the emotion of future utterances. To predict the emotions of the future utterances we propose a novel architecture, Emotion Recognition and Forecasting in Conversation. Our proposed ERFC architecture considers multi modalities, different attributes of emotion, context and the interdependencies of the utterances of the speakers in the conversation. Our intensive experiments on the IEMOCAP dataset have shown the feasibility of the proposed ERFC. This approach can provide a tremendous business value for the applications like call center, where the happiness of customer is utmost important.
Authors: Jay Patel, Hrudayangam Mehta, Jeremy Blackburn
Abstract: Detecting hateful content is a challenging and important problem. Automated tools, like machine-learning models, can help, but they require continuous training to adapt to the ever-changing landscape of social media. In this work, we evaluate eight open-source LLMs' capability to detect antisemitic content, specifically leveraging in-context definition as a policy guideline. We explore various prompting techniques and design a new CoT-like prompt, Guided-CoT. Guided-CoT handles the in-context policy well, increasing performance across all evaluated models, regardless of decoding configuration, model sizes, or reasoning capability. Notably, Llama 3.1 70B outperforms fine-tuned GPT-3.5. Additionally, we examine LLM errors and introduce metrics to quantify semantic divergence in model-generated rationales, revealing notable differences and paradoxical behaviors among LLMs. Our experiments highlight the differences observed across LLMs' utility, explainability, and reliability.
Authors: Hieu Tran, Zonghai Yao, Hong Yu
Abstract: Reinforcement learning improves LLM reasoning, yet sparse delayed reward over long sequences makes token-level credit assignment the key bottleneck. We study the verifiable-reward setting, where the final answer is checkable and multiple responses can be drawn per prompt. Reasoning tasks in math and medical QA align with this setup, where only a few decision tokens significantly impact the outcome. PPO offers token-level advantages with a learned value model, but it is complex to train both the actor and critic models simultaneously, and it is not easily generalizable, as the token-level values from the critic model can make training prone to overfitting. GRPO is critic-free and supports verifiable rewards, but spreads a single sequence-level return across tokens and ignores branching. We introduce \textbf{Prefix-to-Tree (P2T)}, a simple procedure that converts a group of responses into a prefix tree and computes \emph{nonparametric} prefix values \(V(s)\) by aggregating descendant outcomes. Built on P2T, we propose \textbf{TEMPO} (\emph{\textbf{T}ree-\textbf{E}stimated \textbf{M}ean Prefix Value for \textbf{P}olicy \textbf{O}ptimization}), a critic-free algorithm that augments the group-relative outcome signal of GRPO with \emph{branch-gated} temporal-difference corrections derived from the tree. At non-branch tokens, the temporal-difference (TD) term is zero, so TEMPO reduces to GRPO; at branching tokens, it supplies precise token-level credit without a learned value network or extra judges/teachers. On Qwen3-1.7B/4B, TEMPO outperforms PPO and GRPO on in-distribution (MATH, MedQA) and out-of-distribution (GSM-HARD, AMC23, MedMCQA, MMLU-Medical) benchmarks, and reaches higher validation accuracy with roughly the same wall-clock time.
Authors: Saksham Khatwani, He Cheng, Majid Afshar, Dmitriy Dligach, Yanjun Gao
Abstract: Large language models (LLMs) show promise for diagnostic reasoning but often lack reliable, knowledge grounded inference. Knowledge graphs (KGs), such as the Unified Medical Language System (UMLS), offer structured biomedical knowledge that can support trustworthy reasoning. Prior approaches typically integrate KGs via retrieval augmented generation or fine tuning, inserting KG content into prompts rather than enabling structured reasoning. We explore an alternative paradigm: treating the LLM as a reward model of KG reasoning paths, where the model learns to judge whether a candidate path leads to correct diagnosis for a given patient input. This approach is inspired by recent work that leverages reward training to enhance model reasoning abilities, and grounded in computational theory, which suggests that verifying a solution is often easier than generating one from scratch. It also parallels physicians' diagnostic assessment, where they judge which sequences of findings and intermediate conditions most plausibly support a diagnosis. We first systematically evaluate five task formulation for knowledge path judging and eight training paradigm. Second, we test whether the path judging abilities generalize to downstream diagnostic tasks, including diagnosis summarization and medical question answering. Experiments with three open source instruct-tuned LLMs reveal both promise and brittleness: while specific reward optimization and distillation lead to strong path-judging performance, the transferability to downstream tasks remain weak. Our finding provides the first systematic assessment of "reward model style" reasoning over clinical KGs, offering insights into how structured, reward-based supervision influences diagnostic reasoning in GenAI systems for healthcare.
Authors: Pei-Shuo Wang, Jian-Jia Chen, Chun-Che Yang, Chi-Chih Chang, Ning-Chi Huang, Mohamed S. Abdelfattah, Kai-Chiang Wu
Abstract: The immense model sizes of large language models (LLMs) challenge deployment on memory-limited consumer GPUs. Although model compression and parameter offloading are common strategies to address memory limitations, compression can degrade quality, and offloading maintains quality but suffers from slow inference. Speculative decoding presents a promising avenue to accelerate parameter offloading, utilizing a fast draft model to propose multiple draft tokens, which are then verified by the target LLM in parallel with a single forward pass. This method reduces the time-consuming data transfers in forward passes that involve offloaded weight transfers. Existing methods often rely on pretrained weights of the same family, but require additional training to align with custom-trained models. Moreover, approaches that involve draft model training usually yield only modest speedups. This limitation arises from insufficient alignment with the target model, preventing higher token acceptance lengths. To address these challenges and achieve greater speedups, we propose SubSpec, a plug-and-play method to accelerate parameter offloading that is lossless and training-free. SubSpec constructs a highly aligned draft model by generating low-bit quantized substitute layers from offloaded target LLM portions. Additionally, our method shares the remaining GPU-resident layers and the KV-Cache, further reducing memory overhead and enhance alignment. SubSpec achieves a high average acceptance length, delivering 9.1x speedup for Qwen2.5 7B on MT-Bench (8GB VRAM limit) and an average of 12.5x speedup for Qwen2.5 32B on popular generation benchmarks (24GB VRAM limit).
Authors: Chutong Meng, Philipp Koehn
Abstract: We present Speech Vecalign, a parallel speech document alignment method that monotonically aligns speech segment embeddings and does not depend on text transcriptions. Compared to the baseline method Global Mining, a variant of speech mining, Speech Vecalign produces longer speech-to-speech alignments. It also demonstrates greater robustness than Local Mining, another speech mining variant, as it produces less noise. We applied Speech Vecalign to 3,000 hours of unlabeled parallel English-German (En-De) speech documents from VoxPopuli, yielding about 1,000 hours of high-quality alignments. We then trained En-De speech-to-speech translation models on the aligned data. Speech Vecalign improves the En-to-De and De-to-En performance over Global Mining by 0.37 and 0.18 ASR-BLEU, respectively. Moreover, our models match or outperform SpeechMatrix model performance, despite using 8 times fewer raw speech documents.
Authors: Xinlu He, Yiwen Guan, Badrivishal Paurana, Zilin Dai, Jacob Whitehill
Abstract: Most automatic speech processing systems operate in "open loop" mode without user feedback about who said what; yet, human-in-the-loop workflows can potentially enable higher accuracy. We propose an LLM-assisted speaker diarization correction system that lets users fix speaker attribution errors in real time. The pipeline performs streaming ASR and diarization, uses an LLM to deliver concise summaries to the users, and accepts brief verbal feedback that is immediately incorporated without disrupting interactions. Moreover, we develop techniques to make the workflow more effective: First, a split-when-merged (SWM) technique detects and splits multi-speaker segments that the ASR erroneously attributes to just a single speaker. Second, online speaker enrollments are collected based on users' diarization corrections, thus helping to prevent speaker diarization errors from occurring in the future. LLM-driven simulations on the AMI test set indicate that our system substantially reduces DER by 9.92% and speaker confusion error by 44.23%. We further analyze correction efficacy under different settings, including summary vs full transcript display, the number of online enrollments limitation, and correction frequency.
Authors: Minki Hong, Jangho Choi, Jihie Kim
Abstract: Social norms govern culturally appropriate behavior in communication, enabling dialogue systems to produce responses that are not only coherent but also socially acceptable. We present NormGenesis, a multicultural framework for generating and annotating socially grounded dialogues across English, Chinese, and Korean. To model the dynamics of social interaction beyond static norm classification, we propose a novel dialogue type, Violation-to-Resolution (V2R), which models the progression of conversations following norm violations through recognition and socially appropriate repair. To improve pragmatic consistency in underrepresented languages, we implement an exemplar-based iterative refinement early in the dialogue synthesis process. This design introduces alignment with linguistic, emotional, and sociocultural expectations before full dialogue generation begins. Using this framework, we construct a dataset of 10,800 multi-turn dialogues annotated at the turn level for norm adherence, speaker intent, and emotional response. Human and LLM-based evaluations demonstrate that NormGenesis significantly outperforms existing datasets in refinement quality, dialogue naturalness, and generalization performance. We show that models trained on our V2R-augmented data exhibit improved pragmatic competence in ethically sensitive contexts. Our work establishes a new benchmark for culturally adaptive dialogue modeling and provides a scalable methodology for norm-aware generation across linguistically and culturally diverse languages.
Authors: Armin Tourajmehr, Mohammad Reza Modarres, Yadollah Yaghoobzadeh
Abstract: Large language models (LLMs) have demonstrated notable creative abilities in generating literary texts, including poetry and short stories. However, prior research has primarily centered on English, with limited exploration of non-English literary traditions and without standardized methods for assessing creativity. In this paper, we evaluate the capacity of LLMs to generate Persian literary text enriched with culturally relevant expressions. We build a dataset of user-generated Persian literary spanning 20 diverse topics and assess model outputs along four creativity dimensions-originality, fluency, flexibility, and elaboration-by adapting the Torrance Tests of Creative Thinking. To reduce evaluation costs, we adopt an LLM as a judge for automated scoring and validate its reliability against human judgments using intraclass correlation coefficients, observing strong agreement. In addition, we analyze the models' ability to understand and employ four core literary devices: simile, metaphor, hyperbole, and antithesis. Our results highlight both the strengths and limitations of LLMs in Persian literary text generation, underscoring the need for further refinement.
Authors: Oscar J. Ponce-Ponte, David Toro-Tobon, Luis F. Figueroa, Michael Gionfriddo, Megan Branda, Victor M. Montori, Saturnino Luz, Juan P. Brito
Abstract: Shared decision-making (SDM) is necessary to achieve patient-centred care. Currently no methodology exists to automatically measure SDM at scale. This study aimed to develop an automated approach to measure SDM by using language modelling and the conversational alignment (CA) score. A total of 157 video-recorded patient-doctor conversations from a randomized multi-centre trial evaluating SDM decision aids for anticoagulation in atrial fibrillations were transcribed and segmented into 42,559 sentences. Context-response pairs and negative sampling were employed to train deep learning (DL) models and fine-tuned BERT models via the next sentence prediction (NSP) task. Each top-performing model was used to calculate four types of CA scores. A random-effects analysis by clinician, adjusting for age, sex, race, and trial arm, assessed the association between CA scores and SDM outcomes: the Decisional Conflict Scale (DCS) and the Observing Patient Involvement in Decision-Making 12 (OPTION12) scores. p-values were corrected for multiple comparisons with the Benjamini-Hochberg method. Among 157 patients (34% female, mean age 70 SD 10.8), clinicians on average spoke more words than patients (1911 vs 773). The DL model without the stylebook strategy achieved a recall@1 of 0.227, while the fine-tuned BERTbase (110M) achieved the highest recall@1 with 0.640. The AbsMax (18.36 SE7.74 p=0.025) and Max CA (21.02 SE7.63 p=0.012) scores generated with the DL without stylebook were associated with OPTION12. The Max CA score generated with the fine-tuned BERTbase (110M) was associated with the DCS score (-27.61 SE12.63 p=0.037). BERT model sizes did not have an impact the association between CA scores and SDM. This study introduces an automated, scalable methodology to measure SDM in patient-doctor conversations through explainable CA scores, with potential to evaluate SDM strategies at scale.
Authors: Daniel Kaiser, Arnoldo Frigessi, Ali Ramezani-Kebrya, Benjamin Ricaud
Abstract: Current benchmarks for long-context reasoning in Large Language Models (LLMs) often blur critical factors like intrinsic task complexity, distractor interference, and task length. To enable more precise failure analysis, we introduce CogniLoad, a novel synthetic benchmark grounded in Cognitive Load Theory (CLT). CogniLoad generates natural-language logic puzzles with independently tunable parameters that reflect CLT's core dimensions: intrinsic difficulty ($d$) controls intrinsic load; distractor-to-signal ratio ($\rho$) regulates extraneous load; and task length ($N$) serves as an operational proxy for conditions demanding germane load. Evaluating 22 SotA reasoning LLMs, CogniLoad reveals distinct performance sensitivities, identifying task length as a dominant constraint and uncovering varied tolerances to intrinsic complexity and U-shaped responses to distractor ratios. By offering systematic, factorial control over these cognitive load dimensions, CogniLoad provides a reproducible, scalable, and diagnostically rich tool for dissecting LLM reasoning limitations and guiding future model development.
Authors: Zeyu Liu, Souvik Kundu, Lianghao Jiang, Anni Li, Srikanth Ronanki, Sravan Bodapati, Gourav Datta, Peter A. Beerel
Abstract: Although transformer architectures have achieved state-of-the-art performance across diverse domains, their quadratic computational complexity with respect to sequence length remains a significant bottleneck, particularly for latency-sensitive long-context applications. While recent linear-complexity alternatives are increasingly powerful, effectively training them from scratch is still resource-intensive. To overcome these limitations, we propose LAWCAT (Linear Attention with Convolution Across Time), a novel linearization framework designed to efficiently transfer the capabilities of pre-trained transformers into a performant linear attention architecture. LAWCAT integrates causal Conv1D layers to enhance local dependency modeling and employs normalized gated linear attention to improve generalization across varying context lengths. Our comprehensive evaluations demonstrate that, distilling Mistral-7B with only 1K-length sequences yields over 90\% passkey retrieval accuracy up to 22K tokens, significantly extending its effective context window. Similarly, Llama3.2-1B LAWCAT variant achieves competitive performance on S-NIAH 1\&2\&3 tasks (1K-8K context length) and BABILong benchmark (QA2\&QA3, 0K-16K context length), requiring less than 0.1\% pre-training tokens compared with pre-training models. Furthermore, LAWCAT exhibits faster prefill speeds than FlashAttention-2 for sequences exceeding 8K tokens. LAWCAT thus provides an efficient pathway to high-performance, long-context linear models suitable for edge deployment, reducing reliance on extensive long-sequence training data and computational resources.
Authors: Ben Finkelshtein, Silviu Cucerzan, Sujay Kumar Jauhar, Ryen White
Abstract: Large language models (LLMs) are increasingly used for text-rich graph machine learning tasks such as node classification in high-impact domains like fraud detection and recommendation systems. Yet, despite a surge of interest, the field lacks a principled understanding of the capabilities of LLMs in their interaction with graph data. In this work, we conduct a large-scale, controlled evaluation across several key axes of variability to systematically assess the strengths and weaknesses of LLM-based graph reasoning methods in text-based applications. The axes include the LLM-graph interaction mode, comparing prompting, tool-use, and code generation; dataset domains, spanning citation, web-link, e-commerce, and social networks; structural regimes contrasting homophilic and heterophilic graphs; feature characteristics involving both short- and long-text node attributes; and model configurations with varying LLM sizes and reasoning capabilities. We further analyze dependencies by methodically truncating features, deleting edges, and removing labels to quantify reliance on input types. Our findings provide practical and actionable guidance. (1) LLMs as code generators achieve the strongest overall performance on graph data, with especially large gains on long-text or high-degree graphs where prompting quickly exceeds the token budget. (2) All interaction strategies remain effective on heterophilic graphs, challenging the assumption that LLM-based methods collapse under low homophily. (3) Code generation is able to flexibly adapt its reliance between structure, features, or labels to leverage the most informative input type. Together, these findings provide a comprehensive view of the strengths and limitations of current LLM-graph interaction modes and highlight key design principles for future approaches.
Authors: Mohamad Elzohbi, Richard Zhao
Abstract: This paper presents a methodology for inserting phrases in Arabic poems to conform to a specific rhythm using ByT5, a byte-level multilingual transformer-based model. Our work discusses a rule-based grapheme-to-beat transformation tailored for extracting the rhythm from fully diacritized Arabic script. Our approach employs a conditional denoising objective to fine-tune ByT5, where the model reconstructs masked words to match a target rhythm. We adopt a curriculum learning strategy, pre-training on a general Arabic dataset before fine-tuning on poetic dataset, and explore cross-lingual transfer from English to Arabic. Experimental results demonstrate that our models achieve high rhythmic alignment while maintaining semantic coherence. The proposed model has the potential to be used in co-creative applications in the process of composing classical Arabic poems.
Authors: Mo Mu, Dianqiao Lei, Chang Li
Abstract: The widespread adoption of ChatGPT has raised concerns about its misuse, highlighting the need for robust detection of AI-generated text. Current word-level detectors are vulnerable to paraphrasing or simple prompts (PSP), suffer from biases induced by ChatGPT's word-level patterns (CWP) and training data content, degrade on modified text, and often require large models or online LLM interaction. To tackle these issues, we introduce a novel task to detect both original and PSP-modified AI-generated texts, and propose a lightweight framework that classifies texts based on their internal structure, which remains invariant under word-level changes. Our approach encodes sentence embeddings from pre-trained language models and models their relationships via attention. We employ contrastive learning to mitigate embedding biases from autoregressive generation and incorporate a causal graph with counterfactual methods to isolate structural features from topic-related biases. Experiments on two curated datasets, including abstract comparisons and revised life FAQs, validate the effectiveness of our method.
Authors: Jin Young Kim, Ji Won Yoon
Abstract: Recently, inference-time reasoning strategies have further improved the accuracy of large language models (LLMs), but their effectiveness on smaller models remains unclear. Based on the observation that conventional approaches often fail to improve performance in this context, we propose \textbf{C}ycle-\textbf{C}onsistency in \textbf{Q}uestion \textbf{A}nswering (CCQA), a novel reasoning method that can be effectively applied to SLMs. Inspired by cycle consistency, CCQA generates a question from each reasoning path and answer, evaluates each by its similarity to the original question, and then selects the candidate solution with the highest similarity score as the final response. Since conventional SLMs struggle to generate accurate questions from their own reasoning paths and answers, we employ a lightweight Flan-T5 model specialized for question generation to support this process efficiently. From the experimental results, it is verified that CCQA consistently outperforms existing state-of-the-art (SOTA) methods across eight models on mathematical and commonsense reasoning benchmarks. Furthermore, our method establishes a new practical baseline for efficient reasoning in SLMs. Source code can be found at https://github.com/scai-research/ccqa_official.
Authors: Yeongbin Seo, Gayoung Kim, Jaehyung Kim, Jinyoung Yeo
Abstract: As large language models (LLMs) are pretrained on massive web corpora, careful selection of data becomes essential to ensure effective and efficient learning. While perplexity (PPL)-based filtering has shown strong performance, it suffers from drawbacks: substantial time costs and inherent unreliability of the model when handling noisy or out-of-distribution samples. In this work, we propose a simple yet powerful alternative: a prior-based data filtering method that estimates token priors using corpus-level term frequency statistics, inspired by linguistic insights on word roles and lexical density. Our approach filters documents based on the mean and standard deviation of token priors, serving as a fast proxy to PPL while requiring no model inference. Despite its simplicity, the prior-based filter achieves the highest average performance across 20 downstream benchmarks, while reducing time cost by over 1000x compared to PPL-based filtering. We further demonstrate its applicability to symbolic languages such as code and math, and its dynamic adaptability to multilingual corpora without supervision
Authors: Yu Chen, Yifei Han, Long Zhang, Yue Du, Bin Li
Abstract: Fine-tuning large pre-trained models for downstream tasks has become a fundamental approach in natural language processing. Fully fine-tuning all model parameters is computationally expensive and memory-intensive, especially in resource-constrained environments. Existing parameter-efficient fine-tuning methods reduce the number of trainable parameters but typically overlook the varying sensitivity of different model layers and the importance of training data. In this work, we propose TsqLoRA, a novel method that integrates data-quality-driven selection with sensitivity-aware low-rank adaptation, consisted of two main components: a quality-aware sampling mechanism for selecting the most informative training data, and a dynamic rank allocation module that adjusts the rank of each layer based on its sensitivity to parameter updates. The experimental results demonstrate that TsqLoRA improves fine-tuning efficiency while maintaining or even improving performance on a variety of NLP tasks. Our code will be available at https://github.com/Benjamin-Ricky/TsqLoRA.
Authors: Jiarui Jin, Haoyu Wang, Xiang Lan, Jun Li, Gaofeng Cheng, Hongyan Li, Shenda Hong
Abstract: Recent unified models such as GPT-5 have achieved encouraging progress on vision-language tasks. However, these unified models typically fail to correctly understand ECG signals and provide accurate medical diagnoses, nor can they correctly generate ECG signals. To address these limitations, we propose UniECG, the first unified model for ECG capable of concurrently performing evidence-based ECG interpretation and text-conditioned ECG generation tasks. Through a decoupled two-stage training approach, the model first learns evidence-based interpretation skills (ECG-to-Text), and then injects ECG generation capabilities (Text-to-ECG) via latent space alignment. UniECG can autonomously choose to interpret or generate an ECG based on user input, significantly extending the capability boundaries of current ECG models. Our code and checkpoints will be made publicly available at https://github.com/PKUDigitalHealth/UniECG upon acceptance.
Authors: Nishant Balepur, Matthew Shu, Yoo Yeon Sung, Seraphina Goldfarb-Tarrant, Shi Feng, Fumeng Yang, Rachel Rudinger, Jordan Lee Boyd-Graber
Abstract: To assist users in complex tasks, LLMs generate plans: step-by-step instructions towards a goal. While alignment methods aim to ensure LLM plans are helpful, they train (RLHF) or evaluate (ChatbotArena) on what users prefer, assuming this reflects what helps them. We test this with Planorama: an interface where 126 users answer 300 multi-step questions with LLM plans. We get 4388 plan executions and 5584 comparisons to measure plan helpfulness (QA success) and user preferences on plans, and recreate the setup in agents and reward models to see if they simulate or prefer what helps users. We expose: 1) user/model preferences and agent success do not accurately predict which plans help users, so common alignment feedback can misalign with helpfulness; 2) this gap is not due to user-specific preferences, as users are similarly successful when using plans they prefer/disprefer; 3) surface-level cues like brevity and question similarity strongly link to preferences, but such biases fail to predict helpfulness. In all, we argue aligning helpful LLMs needs feedback from real user interactions, not just preferences of what looks helpful, so we discuss the plan NLP researchers can execute to solve this problem.
Authors: Lingwen Deng, Yifei Han, Long Zhang, Yue Du, Bin Li
Abstract: Parameter-Preserving Knowledge Editing (PPKE) enables updating models with new or corrected information without retraining or parameter adjustment. Recent PPKE approaches based on knowledge graphs (KG) to extend knowledge editing (KE) capabilities to multi-hop question answering (MHQA). However, these methods often lack consistency, leading to knowledge contamination, unstable updates, and retrieval behaviors that fail to reflect the intended edits. Such inconsistencies undermine the reliability of PPKE in multi- hop reasoning. We present CAPE-KG, Consistency-Aware Parameter-Preserving Editing with Knowledge Graphs, a novel consistency-aware framework for PPKE on MHQA. CAPE-KG ensures KG construction, update, and retrieval are always aligned with the requirements of the MHQA task, maintaining coherent reasoning over both unedited and edited knowledge. Extensive experiments on the MQuAKE benchmark show accuracy improvements in PPKE performance for MHQA, demonstrating the effectiveness of addressing consistency in PPKE.
Authors: Huanxin Sheng, Xinyi Liu, Hangfeng He, Jieyu Zhao, Jian Kang
Abstract: LLM-as-a-judge has become a promising paradigm for using large language models (LLMs) to evaluate natural language generation (NLG), but the uncertainty of its evaluation remains underexplored. This lack of reliability may limit its deployment in many applications. This work presents the first framework to analyze the uncertainty by offering a prediction interval of LLM-based scoring via conformal prediction. Conformal prediction constructs continuous prediction intervals from a single evaluation run, and we design an ordinal boundary adjustment for discrete rating tasks. We also suggest a midpoint-based score within the interval as a low-bias alternative to raw model score and weighted average. We perform extensive experiments and analysis, which show that conformal prediction can provide valid prediction interval with coverage guarantees. We also explore the usefulness of interval midpoint and judge reprompting for better judgment.
Authors: Yizhe Huang, Yang Liu, Ruiyu Zhao, Xiaolong Zhong, Xingming Yue, Ling Jiang
Abstract: Large Language Model-based agents(LLM-based agents) are increasingly deployed in customer service, yet they often forget across sessions, repeat errors, and lack mechanisms for continual self-improvement. This makes them unreliable in dynamic settings where stability and consistency are critical. To better evaluate these properties, we emphasize two indicators: task success rate as a measure of overall effectiveness, and consistency metrics such as Pass$^k$ to capture reliability across multiple trials. To address the limitations of existing approaches, we propose MemOrb, a lightweight and plug-and-play verbal reinforcement memory layer that distills multi-turn interactions into compact strategy reflections. These reflections are stored in a shared memory bank and retrieved to guide decision-making, without requiring any fine-tuning. Experiments show that MemOrb significantly improves both success rate and stability, achieving up to a 63 percentage-point gain in multi-turn success rate and delivering more consistent performance across repeated trials. Our results demonstrate that structured reflection is a powerful mechanism for enhancing long-term reliability of frozen LLM agents in customer service scenarios.
Authors: Pattara Tipaksorn, Sumonmas Thatphithakkul, Vataya Chunwijitra, Kwanchiva Thangthai
Abstract: We present LOTUSDIS, a publicly available Thai meeting corpus designed to advance far-field conversational ASR. The dataset comprises 114 hours of spontaneous, unscripted dialogue collected in 15-20 minute sessions with three participants, where overlapping speech is frequent and natural. Speech was recorded simultaneously by nine independent single-channel devices spanning six microphone types at distances from 0.12 m to 10 m, preserving the authentic effects of reverberation, noise, and device coloration without relying on microphone arrays. We provide standard train, dev, test splits and release a reproducible baseline system. We benchmarked several Whisper variants under zero-shot and fine-tuned conditions. Off-the-shelf models showed strong degradation with distance, confirming a mismatch between pre-training data and Thai far-field speech. Fine-tuning on LOTUSDIS dramatically improved robustness: a Thai Whisper baseline reduced overall WER from 64.3 to 38.3 and far-field WER from 81.6 to 49.5, with especially large gains on the most distant microphones. These results underscore the importance of distance-diverse training data for robust ASR. The corpus is available under CC-BY-SA 4.0. We also release training and evaluation scripts as a baseline system to promote reproducible research in this field.
Authors: Yunan Wang, Jianxin Li, Ziwei Zhang
Abstract: Dynamic Text-Attribute Graphs (DyTAGs), characterized by time-evolving graph interactions and associated text attributes, are prevalent in real-world applications. Existing methods, such as Graph Neural Networks (GNNs) and Large Language Models (LLMs), mostly focus on static TAGs. Extending these existing methods to DyTAGs is challenging as they largely neglect the recent-global temporal semantics: the recent semantic dependencies among interaction texts and the global semantic evolution of nodes over time. Furthermore, applying LLMs to the abundant and evolving text in DyTAGs faces efficiency issues. To tackle these challenges, we propose Dynamic Global-Recent Adaptive Semantic Processing (DyGRASP), a novel method that leverages LLMs and temporal GNNs to efficiently and effectively reason on DyTAGs. Specifically, we first design a node-centric implicit reasoning method together with a sliding window mechanism to efficiently capture recent temporal semantics. In addition, to capture global semantic dynamics of nodes, we leverage explicit reasoning with tailored prompts and an RNN-like chain structure to infer long-term semantics. Lastly, we intricately integrate the recent and global temporal semantics as well as the dynamic graph structural information using updating and merging layers. Extensive experiments on DyTAG benchmarks demonstrate DyGRASP's superiority, achieving up to 34% improvement in Hit@10 for destination node retrieval task. Besides, DyGRASP exhibits strong generalization across different temporal GNNs and LLMs.
Authors: Julie Kallini, Dan Jurafsky, Christopher Potts, Martijn Bartelds
Abstract: Subword tokenizers trained on multilingual corpora naturally produce overlapping tokens across languages. Does token overlap facilitate cross-lingual transfer or instead introduce interference between languages? Prior work offers mixed evidence, partly due to varied setups and confounders, such as token frequency or subword segmentation granularity. To address this question, we devise a controlled experiment where we train bilingual autoregressive models on multiple language pairs under systematically varied vocabulary overlap settings. Crucially, we explore a new dimension to understanding how overlap affects transfer: the semantic similarity of tokens shared across languages. We first analyze our models' hidden representations and find that overlap of any kind creates embedding spaces that capture cross-lingual semantic relationships, while this effect is much weaker in models with disjoint vocabularies. On XNLI and XQuAD, we find that models with overlap outperform models with disjoint vocabularies, and that transfer performance generally improves as overlap increases. Overall, our findings highlight the advantages of token overlap in multilingual models and show that substantial shared vocabulary remains a beneficial design choice for multilingual tokenizers.
Authors: Yingming Zheng, Hanqi Li, Kai Yu, Lu Chen
Abstract: Large language models (LLMs) have achieved impressive performance across natural language processing (NLP) tasks. As real-world applications increasingly demand longer context windows, continued pretraining and supervised fine-tuning (SFT) on long-context data has become a common approach. While the effects of data length in continued pretraining have been extensively studied, their implications for SFT remain unclear. In this work, we systematically investigate how SFT data length influences LLM behavior on short-context tasks. Counterintuitively, we find that long-context SFT improves short-context performance, contrary to the commonly observed degradation from long-context pretraining. To uncover the underlying mechanisms of this phenomenon, we first decouple and analyze two key components, Multi-Head Attention (MHA) and Feed-Forward Network (FFN), and show that both independently benefit from long-context SFT. We further study their interaction and reveal a knowledge preference bias: long-context SFT promotes contextual knowledge, while short-context SFT favors parametric knowledge, making exclusive reliance on long-context SFT suboptimal. Finally, we demonstrate that hybrid training mitigates this bias, offering explainable guidance for fine-tuning LLMs.
Authors: Wei-Ning Chiu, Yu-Hsiang Wang, Andy Hsiao, Yu-Shiang Huang, Chuan-Ju Wang
Abstract: A multitude of interconnected risk events -- ranging from regulatory changes to geopolitical tensions -- can trigger ripple effects across firms. Identifying inter-firm risk relations is thus crucial for applications like portfolio management and investment strategy. Traditionally, such assessments rely on expert judgment and manual analysis, which are, however, subjective, labor-intensive, and difficult to scale. To address this, we propose a systematic method for extracting inter-firm risk relations using Form 10-K filings -- authoritative, standardized financial documents -- as our data source. Leveraging recent advances in natural language processing, our approach captures implicit and abstract risk connections through unsupervised fine-tuning based on chronological and lexical patterns in the filings. This enables the development of a domain-specific financial encoder with a deeper contextual understanding and introduces a quantitative risk relation score for transparency, interpretable analysis. Extensive experiments demonstrate that our method outperforms strong baselines across multiple evaluation settings.
Authors: Chen Liang, Zhaoqi Huang, Haofen Wang, Fu Chai, Chunying Yu, Huanhuan Wei, Zhengjie Liu, Yanpeng Li, Hongjun Wang, Ruifeng Luo, Xianzhong Zhao
Abstract: Large language models (LLMs), as a novel information technology, are seeing increasing adoption in the Architecture, Engineering, and Construction (AEC) field. They have shown their potential to streamline processes throughout the building lifecycle. However, the robustness and reliability of LLMs in such a specialized and safety-critical domain remain to be evaluated. To address this challenge, this paper establishes AECBench, a comprehensive benchmark designed to quantify the strengths and limitations of current LLMs in the AEC domain. The benchmark defines 23 representative tasks within a five-level cognition-oriented evaluation framework encompassing Knowledge Memorization, Understanding, Reasoning, Calculation, and Application. These tasks were derived from authentic AEC practice, with scope ranging from codes retrieval to specialized documents generation. Subsequently, a 4,800-question dataset encompassing diverse formats, including open-ended questions, was crafted primarily by engineers and validated through a two-round expert review. Furthermore, an LLM-as-a-Judge approach was introduced to provide a scalable and consistent methodology for evaluating complex, long-form responses leveraging expert-derived rubrics. Through the evaluation of nine LLMs, a clear performance decline across five cognitive levels was revealed. Despite demonstrating proficiency in foundational tasks at the Knowledge Memorization and Understanding levels, the models showed significant performance deficits, particularly in interpreting knowledge from tables in building codes, executing complex reasoning and calculation, and generating domain-specific documents. Consequently, this study lays the groundwork for future research and development aimed at the robust and reliable integration of LLMs into safety-critical engineering practices.
Authors: Sabri Boughorbel, Fahim Dalvi, Nadir Durrani, Majd Hawasly
Abstract: As fine-tuning becomes the dominant paradigm for improving large language models (LLMs), understanding what changes during this process is increasingly important. Traditional benchmarking often fails to explain why one model outperforms another. In this work, we use model diffing, a mechanistic interpretability approach, to analyze the specific capability differences between Gemma-2-9b-it and a SimPO-enhanced variant. Using crosscoders, we identify and categorize latent representations that differentiate the two models. We find that SimPO acquired latent concepts predominantly enhance safety mechanisms (+32.8%), multilingual capabilities (+43.8%), and instruction-following (+151.7%), while its additional training also reduces emphasis on model self-reference (-44.1%) and hallucination management (-68.5%). Our analysis shows that model diffing can yield fine-grained insights beyond leaderboard metrics, attributing performance gaps to concrete mechanistic capabilities. This approach offers a transparent and targeted framework for comparing LLMs.
Authors: Liting Zhang, Shiwan Zhao, Aobo Kong, Qicheng Li
Abstract: Keyphrase extraction is a fundamental task in natural language processing. However, existing unsupervised prompt-based methods for Large Language Models (LLMs) often rely on single-stage inference pipelines with uniform prompting, regardless of document length or LLM backbone. Such one-size-fits-all designs hinder the full exploitation of LLMs' reasoning and generation capabilities, especially given the complexity of keyphrase extraction across diverse scenarios. To address these challenges, we propose MAPEX, the first framework that introduces multi-agent collaboration into keyphrase extraction. MAPEX coordinates LLM-based agents through modules for expert recruitment, candidate extraction, topic guidance, knowledge augmentation, and post-processing. A dual-path strategy dynamically adapts to document length: knowledge-driven extraction for short texts and topic-guided extraction for long texts. Extensive experiments on six benchmark datasets across three different LLMs demonstrate its strong generalization and universality, outperforming the state-of-the-art unsupervised method by 2.44\% and standard LLM baselines by 4.01\% in F1@5 on average. Code is available at https://github.com/NKU-LITI/MAPEX.
Authors: Damian Stachura, Joanna Konieczna, Artur Nowak
Abstract: Open-weight versions of large language models (LLMs) are rapidly advancing, with state-of-the-art models like DeepSeek-V3 now performing comparably to proprietary LLMs. This progression raises the question of whether small open-weight LLMs are capable of effectively replacing larger closed-source models. We are particularly interested in the context of biomedical question-answering, a domain we explored by participating in Task 13B Phase B of the BioASQ challenge. In this work, we compare several open-weight models against top-performing systems such as GPT-4o, GPT-4.1, Claude 3.5 Sonnet, and Claude 3.7 Sonnet. To enhance question answering capabilities, we use various techniques including retrieving the most relevant snippets based on embedding distance, in-context learning, and structured outputs. For certain submissions, we utilize ensemble approaches to leverage the diverse outputs generated by different models for exact-answer questions. Our results demonstrate that open-weight LLMs are comparable to proprietary ones. In some instances, open-weight LLMs even surpassed their closed counterparts, particularly when ensembling strategies were applied. All code is publicly available at https://github.com/evidenceprime/BioASQ-13b.
Authors: Luyan Zhang, Xinyu Xie
Abstract: With the rapid advancement of large language model technology, there is growing interest in whether multi-feature approaches can significantly improve AI text detection beyond what single neural models achieve. While intuition suggests that combining semantic, syntactic, and statistical features should provide complementary signals, this assumption has not been rigorously tested with modern LLM-generated text. This paper provides a systematic empirical investigation of multi-hierarchical feature integration for AI text detection, specifically testing whether the computational overhead of combining multiple feature types is justified by performance gains. We implement MHFD (Multi-Hierarchical Feature Detection), integrating DeBERTa-based semantic analysis, syntactic parsing, and statistical probability features through adaptive fusion. Our investigation reveals important negative results: despite theoretical expectations, multi-feature integration provides minimal benefits (0.4-0.5% improvement) while incurring substantial computational costs (4.2x overhead), suggesting that modern neural language models may already capture most relevant detection signals efficiently. Experimental results on multiple benchmark datasets demonstrate that the MHFD method achieves 89.7% accuracy in in-domain detection and maintains 84.2% stable performance in cross-domain detection, showing modest improvements of 0.4-2.6% over existing methods.
Authors: Advik Raj Basani, Pin-Yu Chen
Abstract: Detecting AI-generated text is an increasing necessity to combat misuse of LLMs in education, business compliance, journalism, and social media, where synthetic fluency can mask misinformation or deception. While prior detectors often rely on token-level likelihoods or opaque black-box classifiers, these approaches struggle against high-quality generations and offer little interpretability. In this work, we propose DivEye, a novel detection framework that captures how unpredictability fluctuates across a text using surprisal-based features. Motivated by the observation that human-authored text exhibits richer variability in lexical and structural unpredictability than LLM outputs, DivEye captures this signal through a set of interpretable statistical features. Our method outperforms existing zero-shot detectors by up to 33.2% and achieves competitive performance with fine-tuned baselines across multiple benchmarks. DivEye is robust to paraphrasing and adversarial attacks, generalizes well across domains and models, and improves the performance of existing detectors by up to 18.7% when used as an auxiliary signal. Beyond detection, DivEye provides interpretable insights into why a text is flagged, pointing to rhythmic unpredictability as a powerful and underexplored signal for LLM detection.
Authors: Nicholas Popovi\v{c}, Michael F\"arber
Abstract: Recent works in Natural Language Inference (NLI) and related tasks, such as automated fact-checking, employ atomic fact decomposition to enhance interpretability and robustness. For this, existing methods rely on resource-intensive generative large language models (LLMs) to perform decomposition. We propose JEDI, an encoder-only architecture that jointly performs extractive atomic fact decomposition and interpretable inference without requiring generative models during inference. To facilitate training, we produce a large corpus of synthetic rationales covering multiple NLI benchmarks. Experimental results demonstrate that JEDI achieves competitive accuracy in distribution and significantly improves robustness out of distribution and in adversarial settings over models based solely on extractive rationale supervision. Our findings show that interpretability and robust generalization in NLI can be realized using encoder-only architectures and synthetic rationales. Code and data available at https://jedi.nicpopovic.com
Authors: Abderrahmane Issam, Yusuf Can Semerci, Jan Scholtes, Gerasimos Spanakis
Abstract: End-to-End Speech Translation (E2E-ST) is the task of translating source speech directly into target text bypassing the intermediate transcription step. The representation discrepancy between the speech and text modalities has motivated research on what is known as bridging the modality gap. State-of-the-art methods addressed this by aligning speech and text representations on the word or token level. Unfortunately, this requires an alignment tool that is not available for all languages. Although this issue has been addressed by aligning speech and text embeddings using nearest-neighbor similarity search, it does not lead to accurate alignments. In this work, we adapt Dynamic Time Warping (DTW) for aligning speech and text embeddings during training. Our experiments demonstrate the effectiveness of our method in bridging the modality gap in E2E-ST. Compared to previous work, our method produces more accurate alignments and achieves comparable E2E-ST results while being significantly faster. Furthermore, our method outperforms previous work in low resource settings on 5 out of 6 language directions.
Authors: Shaomu Tan, Ryosuke Mitani, Ritvik Choudhary, Toshiyuki Sekiya
Abstract: Scaling model parameters has become the de facto strategy for improving NLP systems, but it comes with substantial computational costs. Test-Time Scaling (TTS) offers an alternative by allocating more computation at inference: generating multiple candidates and selecting the best. While effective in tasks such as mathematical reasoning, TTS has not been systematically explored for machine translation (MT). In this paper, we present the first systematic study of TTS for MT, investigating a simple but practical best-of-N framework on WMT24 benchmarks. Our experiments cover six high-resource and one low-resource language pairs, five model sizes (3B-72B), and various TTS compute budget (N up to 1024). Our results show that a) For high-resource languages, TTS generally improves translation quality according to multiple neural MT evaluation metrics, and our human evaluation confirms these gains; b) Augmenting smaller models with large $N$ can match or surpass larger models at $N{=}1$ with more compute cost; c) Under fixed compute budgets, larger models are typically more efficient, and TTS can degrade quality due to metric blind spots in low-resource cases.
Authors: Chiara Alzetta, Serena Auriemma, Alessandro Bondielli, Luca Dini, Chiara Fazzone, Alessio Miaschi, Martina Miliani, Marta Sartor
Abstract: Over the past decade, Computational Linguistics (CL) and Natural Language Processing (NLP) have evolved rapidly, especially with the advent of Transformer-based Large Language Models (LLMs). This shift has transformed research goals and priorities, from Lexical and Semantic Resources to Language Modelling and Multimodality. In this study, we track the research trends of the Italian CL and NLP community through an analysis of the contributions to CLiC-it, arguably the leading Italian conference in the field. We compile the proceedings from the first 10 editions of the CLiC-it conference (from 2014 to 2024) into the CLiC-it Corpus, providing a comprehensive analysis of both its metadata, including author provenance, gender, affiliations, and more, as well as the content of the papers themselves, which address various topics. Our goal is to provide the Italian and international research communities with valuable insights into emerging trends and key developments over time, supporting informed decisions and future directions in the field.
Authors: Alireza Salemi, Cheng Li, Mingyang Zhang, Qiaozhu Mei, Zhuowan Li, Spurthi Amba Hombaiah, Weize Kong, Tao Chen, Hamed Zamani, Michael Bendersky
Abstract: Personalization is essential for adapting question answering (QA) systems to user-specific information needs, thereby improving both accuracy and user satisfaction. However, personalized QA remains relatively underexplored due to challenges such as inferring preferences from long, noisy, and implicit contexts, and generating responses that are simultaneously correct, contextually appropriate, and aligned with user expectations and background knowledge. To address these challenges, we propose Pathways of Thoughts (PoT), an inference-stage method that applies to any large language model (LLM) without requiring task-specific fine-tuning. The approach models the reasoning of an LLM as an iterative decision process, where the model dynamically selects among cognitive operations such as reasoning, revision, personalization, and clarification. This enables exploration of multiple reasoning trajectories, producing diverse candidate responses that capture different perspectives. PoT then aggregates and reweights these candidates according to inferred user preferences, yielding a final personalized response that benefits from the complementary strengths of diverse reasoning paths. Experiments on the LaMP-QA benchmark for personalized QA show that PoT consistently outperforms competitive baselines, achieving up to a 13.1% relative improvement. Human evaluation corroborates these results, with annotators preferring outputs from PoT in 66% of cases and reporting ties in only 15% of cases.
Authors: Hiram Ring
Abstract: A repeated claim in linguistics is that the majority of linguistic utterances are unique. For example, Pinker (1994: 10), summarizing an argument by Noam Chomsky, states that "virtually every sentence that a person utters or understands is a brand-new combination of words, appearing for the first time in the history of the universe." With the increased availability of large corpora, this is a claim that can be empirically investigated. The current paper addresses the question by using the NLTK Python library to parse corpora of different genres, providing counts of exact string matches in each. Results show that while completely unique sentences are often the majority of corpora, this is highly constrained by genre, and that duplicate sentences are not an insignificant part of any individual corpus.
Authors: Timur Turatali, Anton Alekseev, Gulira Jumalieva, Gulnara Kabaeva, Sergey Nikolenko
Abstract: We introduce KyrgyzNER, the first manually annotated named entity recognition dataset for the Kyrgyz language. Comprising 1,499 news articles from the 24.KG news portal, the dataset contains 10,900 sentences and 39,075 entity mentions across 27 named entity classes. We show our annotation scheme, discuss the challenges encountered in the annotation process, and present the descriptive statistics. We also evaluate several named entity recognition models, including traditional sequence labeling approaches based on conditional random fields and state-of-the-art multilingual transformer-based models fine-tuned on our dataset. While all models show difficulties with rare entity categories, models such as the multilingual RoBERTa variant pretrained on a large corpus across many languages achieve a promising balance between precision and recall. These findings emphasize both the challenges and opportunities of using multilingual pretrained models for processing languages with limited resources. Although the multilingual RoBERTa model performed best, other multilingual models yielded comparable results. This suggests that future work exploring more granular annotation schemes may offer deeper insights for Kyrgyz language processing pipelines evaluation.
Authors: Kun Zhu, Lizi Liao, Yuxuan Gu, Lei Huang, Xiaocheng Feng, Bing Qin
Abstract: The rapid growth of scientific literature demands efficient methods to organize and synthesize research findings. Existing taxonomy construction methods, leveraging unsupervised clustering or direct prompting of large language models (LLMs), often lack coherence and granularity. We propose a novel context-aware hierarchical taxonomy generation framework that integrates LLM-guided multi-aspect encoding with dynamic clustering. Our method leverages LLMs to identify key aspects of each paper (e.g., methodology, dataset, evaluation) and generates aspect-specific paper summaries, which are then encoded and clustered along each aspect to form a coherent hierarchy. In addition, we introduce a new evaluation benchmark of 156 expert-crafted taxonomies encompassing 11.6k papers, providing the first naturally annotated dataset for this task. Experimental results demonstrate that our method significantly outperforms prior approaches, achieving state-of-the-art performance in taxonomy coherence, granularity, and interpretability.
Authors: Alejandro Cuevas, Saloni Dash, Bharat Kumar Nayak, Dan Vann, Madeleine I. G. Daepp
Abstract: Disinformation is among the top risks of generative artificial intelligence (AI) misuse. Global adoption of generative AI necessitates red-teaming evaluations (i.e., systematic adversarial probing) that are robust across diverse languages and cultures, but red-teaming datasets are commonly US- and English-centric. To address this gap, we propose "anecdoctoring", a novel red-teaming approach that automatically generates adversarial prompts across languages and cultures. We collect misinformation claims from fact-checking websites in three languages (English, Spanish, and Hindi) and two geographies (US and India). We then cluster individual claims into broader narratives and characterize the resulting clusters with knowledge graphs, with which we augment an attacker LLM. Our method produces higher attack success rates and offers interpretability benefits relative to few-shot prompting. Results underscore the need for disinformation mitigations that scale globally and are grounded in real-world adversarial misuse.
Authors: Chantal Shaib, Tuhin Chakrabarty, Diego Garcia-Olano, Byron C. Wallace
Abstract: AI "slop" is an increasingly popular term used to describe low-quality AI-generated text, but there is currently no agreed upon definition of this term nor a means to measure its occurrence. In this work, we develop a taxonomy of "slop" through interviews with experts in NLP, writing, and philosophy, and propose a set of interpretable dimensions for its assessment in text. Through span-level annotation, we find that binary "slop" judgments are (somewhat) subjective, but such determinations nonetheless correlate with latent dimensions such as coherence and relevance. Our framework can be used to evaluate AI-generated text in both detection and binary preference tasks, potentially offering new insights into the linguistic and stylistic factors that contribute to quality judgments.
Authors: Natasha Butt, Ariel Kwiatkowski, Ismail Labiad, Julia Kempe, Yann Ollivier
Abstract: The use of continuous instead of discrete tokens during the Chain-of-Thought (CoT) phase of reasoning LLMs has garnered attention recently, based on the intuition that a continuous mixture of discrete tokens could simulate a superposition of several reasoning paths simultaneously. Theoretical results have formally proven that continuous tokens have much greater expressivity and can solve specific problems more efficiently. However, practical use of continuous tokens has been limited by strong training difficulties: previous works either just use continuous tokens at inference time on a pre-trained discrete-token model, or must distill the continuous CoT from ground-truth discrete CoTs and face computational costs that limit the CoT to very few tokens. This is the first work introducing a scalable method to learn continuous CoTs via reinforcement learning (RL), without distilling from reference discrete CoTs. We use "soft" tokens: mixtures of tokens together with noise on the input embedding to provide RL exploration. Computational overhead is minimal, enabling us to learn continuous CoTs with hundreds of tokens. On math reasoning benchmarks with Llama and Qwen models up to 8B, training with continuous CoTs match discrete-token CoTs for pass@1 and surpass them for pass@32, showing greater CoT diversity. In systematic comparisons, the best-performing scenario is to train with continuous CoT tokens then use discrete tokens for inference, meaning the "soft" models can be deployed in a standard way. Finally, we show continuous CoT RL training better preserves the predictions of the base model on out-of-domain tasks, thus providing a softer touch to the base model.
Authors: Xiaoqian Liu, Ke Wang, Yuchuan Wu, Fei Huang, Yongbin Li, Junge Zhang, Jianbin Jiao
Abstract: Large language models (LLMs) are increasingly trained with reinforcement learning (RL) as autonomous agents that reason and act over long horizons in interactive environments. However, sparse and sometimes unverifiable rewards make temporal credit assignment extremely challenging. Recent work attempts to integrate process supervision into agent learning but suffers from biased annotation, reward hacking, high-variance from overly fine-grained signals or failtures when state overlap is rare. We therefore introduce Online Process Reward Learning (OPRL), a general credit-assignment strategy for agentic RL that integrates seamlessly with standard on-policy algorithms without relying on additional rollouts or explicit step labels. In OPRL, we optimize an implicit process reward model (PRM) alternately with the agent's policy to transform trajectory preferences into implicit step rewards through a trajectory-based DPO objective. These step rewards are then used to compute step-level advantages, which are combined with episode-level advantages from outcome rewards for policy update, creating a self-reinforcing loop. Theoretical findings guarantee that the learned step rewards are consistent with trajectory preferences and act as potential-based shaping rewards, providing bounded gradients to stabilize training. Empirically, we evaluate OPRL on three distinct agent benmarks, including WebShop and VisualSokoban, as well as open-ended social interactions with unverfiable rewards in SOTOPIA. Crucially, OPRL shows superior performance over frontier LLMs and strong RL baselines across domains, achieving state-of-the-art results with higher sample-efficiency and lower variance during training. Further analysis also demonstrates the efficient exploration by OPRL using fewer actions, underscoring its potential for agentic learning in real-world scenarios.
Authors: Zheyuan Liu, Zhangchen Xu, Guangyao Dou, Xiangchi Yuan, Zhaoxuan Tan, Radha Poovendran, Meng Jiang
Abstract: Multimodal Large Language Models (MLLMs) are increasingly deployed in real-world applications, yet their ability to make context-aware safety decisions remains limited. Existing methods often fail to balance oversensitivity (unjustified refusals of benign queries) and undersensitivity (missed detection of visually grounded risks), leaving a persistent gap in safety alignment. To address this issue, we introduce Safety-aware Contrastive Decoding (SafeCoDe), a lightweight and model-agnostic decoding framework that dynamically adjusts token generation based on multimodal context. SafeCoDe operates in two stages: (1) a contrastive decoding mechanism that highlights tokens sensitive to visual context by contrasting real and Gaussian-noised images, and (2) a global-aware token modulation strategy that integrates scene-level reasoning with token-level adjustment to adapt refusals according to the predicted safety verdict. Extensive experiments across diverse MLLM architectures and safety benchmarks, covering undersensitivity, oversensitivity, and general safety evaluations, show that SafeCoDe consistently improves context-sensitive refusal behaviors while preserving model helpfulness.
Authors: Tariq Abdul-Quddoos, Xishuang Dong, Lijun Qian
Abstract: Attention-based models have become the leading approach in modeling medical language for Natural Language Processing (NLP) in clinical notes. These models outperform traditional techniques by effectively capturing contextual rep- resentations of language. In this research a comparative analysis is done amongst pre- trained attention based models namely Bert Base, BioBert, two variations of Bio+Clinical Bert, RoBerta, and Clinical Long- former on task related to Electronic Health Record (EHR) information extraction. The tasks from Track 1 of Harvard Medical School's 2022 National Clinical NLP Challenges (n2c2) are considered for this comparison, with the Contextualized Medication Event Dataset (CMED) given for these task. CMED is a dataset of unstructured EHRs and annotated notes that contain task relevant information about the EHRs. The goal of the challenge is to develop effective solutions for extracting contextual information related to patient medication events from EHRs using data driven methods. Each pre-trained model is fine-tuned and applied on CMED to perform medication extraction, medical event detection, and multi-dimensional medication event context classification. Pro- cessing methods are also detailed for breaking down EHRs for compatibility with the applied models. Performance analysis has been carried out using a script based on constructing medical terms from the evaluation portion of CMED with metrics including recall, precision, and F1-Score. The results demonstrate that models pre-trained on clinical data are more effective in detecting medication and medication events, but Bert Base, pre- trained on general domain data showed to be the most effective for classifying the context of events related to medications.
Authors: Gabriele Berton, Jayakrishnan Unnikrishnan, Son Tran, Mubarak Shah
Abstract: Large Language Models (LLMs) face significant computational challenges when processing long contexts due to the quadratic complexity of self-attention. While soft context compression methods, which map input text to smaller latent representations, have shown promise, their real-world adoption is limited. Existing techniques typically compress the context as a single unit, which leads to quadratic compression complexity and an inability to reuse computations across queries with overlapping contexts. In this work, we introduce CompLLM, a soft compression technique designed for practical deployment. Instead of processing the context holistically, CompLLM divides it into segments and compresses each one independently. This simple design choice yields three critical properties: efficiency, as the compression step scales linearly with the context length; scalability, enabling models trained on short sequences (e.g., 1k tokens) to generalize to contexts of 100k tokens; and reusability, allowing compressed segments to be cached and reused across different queries. Our experiments show that with a 2x compression rate, at high context lengths CompLLM speeds up Time To First Token (TTFT) by up to 4x and reduces the KV cache size by 50%. Furthermore, CompLLM achieves performance comparable to that obtained with the uncompressed context, and even surpasses it on very long sequences, demonstrating its effectiveness and practical utility.
Authors: Siheng Li, Kejiao Li, Zenan Xu, Guanhua Huang, Evander Yang, Kun Li, Haoyuan Wu, Jiajia Wu, Zihao Zheng, Chenchen Zhang, Kun Shi, Kyrierl Deng, Qi Yi, Ruibin Xiong, Tingqiang Xu, Yuhao Jiang, Jianfeng Yan, Yuyuan Zeng, Guanghui Xu, Jinbao Xue, Zhijiang Xu, Zheng Fang, Shuai Li, Qibin Liu, Xiaoxue Li, Zhuoyu Li, Yangyu Tao, Fei Gao, Cheng Jiang, Bo Chao Wang, Kai Liu, Jianchen Zhu, Wai Lam, Wayyt Wang, Bo Zhou, Di Wang
Abstract: The growing disparity between the exponential scaling of computational resources and the finite growth of high-quality text data now constrains conventional scaling approaches for large language models (LLMs). To address this challenge, we introduce Reinforcement Learning on Pre-Training data (RLPT), a new training-time scaling paradigm for optimizing LLMs. In contrast to prior approaches that scale training primarily through supervised learning, RLPT enables the policy to autonomously explore meaningful trajectories to learn from pre-training data and improve its capability through reinforcement learning (RL). While existing RL strategies such as reinforcement learning from human feedback (RLHF) and reinforcement learning with verifiable rewards (RLVR) rely on human annotation for reward construction, RLPT eliminates this dependency by deriving reward signals directly from pre-training data. Specifically, it adopts a next-segment reasoning objective, rewarding the policy for accurately predicting subsequent text segments conditioned on the preceding context. This formulation allows RL to be scaled on pre-training data, encouraging the exploration of richer trajectories across broader contexts and thereby fostering more generalizable reasoning skills. Extensive experiments on both general-domain and mathematical reasoning benchmarks across multiple models validate the effectiveness of RLPT. For example, when applied to Qwen3-4B-Base, RLPT yields absolute improvements of $3.0$, $5.1$, $8.1$, $6.0$, $6.6$, and $5.3$ on MMLU, MMLU-Pro, GPQA-Diamond, KOR-Bench, AIME24, and AIME25, respectively. The results further demonstrate favorable scaling behavior, suggesting strong potential for continued gains with more compute. In addition, RLPT provides a solid foundation, extending the reasoning boundaries of LLMs and enhancing RLVR performance.
Authors: Nitesh Kumar, Usashi Chatterjee, Steven Schockaert
Abstract: Conceptual spaces represent entities and concepts using cognitively meaningful dimensions, typically referring to perceptual features. Such representations are widely used in cognitive science and have the potential to serve as a cornerstone for explainable AI. Unfortunately, they have proven notoriously difficult to learn, although recent LLMs appear to capture the required perceptual features to a remarkable extent. Nonetheless, practical methods for extracting the corresponding conceptual spaces are currently still lacking. While various methods exist for extracting embeddings from LLMs, extracting conceptual spaces also requires us to encode the underlying features. In this paper, we propose a strategy in which features (e.g. sweetness) are encoded by embedding the description of a corresponding prototype (e.g. a very sweet food). To improve this strategy, we fine-tune the LLM to align the prototype embeddings with the corresponding conceptual space dimensions. Our empirical analysis finds this approach to be highly effective.
Authors: Erik Bo\v{z}\'ik, Marek \v{S}uppa
Abstract: Automatic Speech Recognition (ASR) for low-resource languages like Slovak is hindered by the scarcity of training data. To address this, we introduce SloPalSpeech, a new, large-scale Slovak ASR dataset containing 2,806 hours of speech from parliamentary proceedings. We developed a robust processing pipeline to align and segment long-form recordings into clean, 30-second audio-transcript pairs suitable for model training. We use this dataset to fine-tune several OpenAI Whisper models (small, medium, large-v3, and large-v3-turbo), achieving significant Word Error Rate (WER) reductions on standard Slovak benchmarks like Common Voice and FLEURS. For instance, the fine-tuned Whisper-small model's WER dropped by up to 70\%, approaching the baseline performance of the much larger Whisper-large-v3 model. To foster future research in low-resource speech recognition, we publicly release the complete SloPalSpeech dataset, the fully segmented transcripts (60 million words), and all our fine-tuned models.
Authors: Abdou Karim Kandji, Fr\'ed\'eric Precioso, Cheikh Ba, Samba Ndiaye, Augustin Ndione
Abstract: Intent classification models have made a lot of progress in recent years. However, previous studies primarily focus on high-resource languages datasets, which results in a gap for low-resource languages and for regions with a high rate of illiterate people where languages are more spoken than read or written. This is the case in Senegal, for example, where Wolof is spoken by around 90\% of the population, with an illiteracy rate of 42\% for the country. Wolof is actually spoken by more than 10 million people in West African region. To tackle such limitations, we release a Wolof Intent Classification Dataset (WolBanking77), for academic research in intent classification. WolBanking77 currently contains 9,791 text sentences in the banking domain and more than 4 hours of spoken sentences. Experiments on various baselines are conducted in this work, including text and voice state-of-the-art models. The results are very promising on this current dataset. This paper also provides detailed analyses of the contents of the data. We report baseline f1-score and word error rate metrics respectively on NLP and ASR models trained on WolBanking77 dataset and also comparisons between models. We plan to share and conduct dataset maintenance, updates and to release open-source code.
Authors: Arijit Maji, Raghvendra Kumar, Akash Ghosh, Anushka, Nemil Shah, Abhilekh Borah, Vanshika Shah, Nishant Mishra, Sriparna Saha
Abstract: We introduce DRISHTIKON, a first-of-its-kind multimodal and multilingual benchmark centered exclusively on Indian culture, designed to evaluate the cultural understanding of generative AI systems. Unlike existing benchmarks with a generic or global scope, DRISHTIKON offers deep, fine-grained coverage across India's diverse regions, spanning 15 languages, covering all states and union territories, and incorporating over 64,000 aligned text-image pairs. The dataset captures rich cultural themes including festivals, attire, cuisines, art forms, and historical heritage amongst many more. We evaluate a wide range of vision-language models (VLMs), including open-source small and large models, proprietary systems, reasoning-specialized VLMs, and Indic-focused models, across zero-shot and chain-of-thought settings. Our results expose key limitations in current models' ability to reason over culturally grounded, multimodal inputs, particularly for low-resource languages and less-documented traditions. DRISHTIKON fills a vital gap in inclusive AI research, offering a robust testbed to advance culturally aware, multimodally competent language technologies.
Authors: Jiaqi Weng, Han Zheng, Hanyu Zhang, Qinqin He, Jialing Tao, Hui Xue, Zhixuan Chu, Xiting Wang
Abstract: Increasing deployment of large language models (LLMs) in real-world applications raises significant safety concerns. Most existing safety research focuses on evaluating LLM outputs or specific safety tasks, limiting their ability to ad- dress broader, undefined risks. Sparse Autoencoders (SAEs) facilitate interpretability research to clarify model behavior by explaining single-meaning atomic features decomposed from entangled signals. jHowever, prior applications on SAEs do not interpret features with fine-grained safety-related con- cepts, thus inadequately addressing safety-critical behaviors, such as generating toxic responses and violating safety regu- lations. For rigorous safety analysis, we must extract a rich and diverse set of safety-relevant features that effectively capture these high-risk behaviors, yet face two challenges: identifying SAEs with the greatest potential for generating safety concept-specific neurons, and the prohibitively high cost of detailed feature explanation. In this paper, we pro- pose Safe-SAIL, a framework for interpreting SAE features within LLMs to advance mechanistic understanding in safety domains. Our approach systematically identifies SAE with best concept-specific interpretability, explains safety-related neurons, and introduces efficient strategies to scale up the in- terpretation process. We will release a comprehensive toolkit including SAE checkpoints and human-readable neuron ex- planations, which supports empirical analysis of safety risks to promote research on LLM safety.
Authors: Hengbo Xiao, Jingyuan Fan, Xin Tong, Jingzhao Zhang, Chao Lu, Guannan He
Abstract: Complex systems typically rely on high-precision numerical computation to support decisions, but current large language models (LLMs) cannot yet incorporate such computations as an intrinsic and interpretable capability with existing architectures. Mainstream multi-agent approaches can leverage external experts, but inevitably introduce communication overhead and suffer from inefficient multimodal emergent capability and limited scalability. To this end, we propose PiMoE (Physically-isolated Mixture of Experts), a training and inference architecture for integrating computation and reasoning. Instead of the workflow paradigm of tool invocation, PiMoE endogenously integrates computational capabilities into neural networks after separately training experts, a text-to-computation module, and a router. At inference, the router directs computation and reasoning at the token level, thereby enabling iterative alternation within a single chain of thought. We evaluate PiMoE on two reasoning-computation tasks against LLM finetuning and the multi-agent system approaches. Results show that the PiMoE architecture achieves not only higher accuracy than directly finetuning LLMs but also significant improvements in response latency, token usage, and GPU energy consumption compared with mainstream multi-agent approaches. PiMoE offers an efficient, interpretable, and scalable paradigm for next-generation scientific or industrial intelligent systems.
Authors: Hongyi Luo, Qing Cheng, Daniel Matos, Hari Krishna Gadi, Yanfeng Zhang, Lu Liu, Yongliang Wang, Niclas Zeller, Daniel Cremers, Liqiu Meng
Abstract: Humans can interpret geospatial information through natural language, while the geospatial cognition capabilities of Large Language Models (LLMs) remain underexplored. Prior research in this domain has been constrained by non-quantifiable metrics, limited evaluation datasets and unclear research hierarchies. Therefore, we propose a large-scale benchmark and conduct a comprehensive evaluation of the geospatial route cognition of LLMs. We create a large-scale evaluation dataset comprised of 36000 routes from 12 metropolises worldwide. Then, we introduce PathBuilder, a novel tool for converting natural language instructions into navigation routes, and vice versa, bridging the gap between geospatial information and natural language. Finally, we propose a new evaluation framework and metrics to rigorously assess 11 state-of-the-art (SOTA) LLMs on the task of route reversal. The benchmark reveals that LLMs exhibit limitation to reverse routes: most reverse routes neither return to the starting point nor are similar to the optimal route. Additionally, LLMs face challenges such as low robustness in route generation and high confidence for their incorrect answers. Code\ \&\ Data available here: \href{https://github.com/bghjmn32/EMNLP2025_Turnback}{TurnBack.}
Authors: Khalil Hennara, Muhammad Hreden, Mohamed Motasim Hamed, Ahmad Bastati, Zeina Aldallal, Sara Chrouf, Safwan AlModhayan
Abstract: Arabic document OCR remains a challenging task due to the language's cursive script, diverse fonts, diacritics, and right-to-left orientation. While modern Multimodal Large Language Models (MLLMs) have advanced document understanding for high-resource languages, their performance on Arabic remains limited. In this work, we introduce Baseer, a vision-language model fine- tuned specifically for Arabic document OCR. Leveraging a large-scale dataset combining synthetic and real-world documents, Baseer is trained using a decoder-only fine-tuning strategy to adapt a pre-trained MLLM while preserving general visual features. We also present Misraj-DocOCR, a high-quality, expert-verified benchmark designed for rigorous evaluation of Arabic OCR systems. Our experiments show that Baseer significantly outperforms existing open-source and commercial solutions, achieving a WER of 0.25 and establishing a new state-of-the-art in the domain of Arabic document OCR. Our results highlight the benefits of domain-specific adaptation of general-purpose MLLMs and establish a strong baseline for high-accuracy OCR on morphologically rich languages like Arabic.
Authors: Yu Ti Huang
Abstract: Conversational agents must translate egocentric utterances (e.g., "on my right") into allocentric orientations (N/E/S/W). This challenge is particularly critical in indoor or complex facilities where GPS signals are weak and detailed maps are unavailable. While chain-of-thought (CoT) prompting has advanced reasoning in language and vision tasks, its application to multimodal spatial orientation remains underexplored. We introduce Conversational Orientation Reasoning (COR), a new benchmark designed for Traditional Chinese conversational navigation projected from real-world environments, addressing egocentric-to-allocentric reasoning in non-English and ASR-transcribed scenarios. We propose a multimodal chain-of-thought (MCoT) framework, which integrates ASR-transcribed speech with landmark coordinates through a structured three-step reasoning process: (1) extracting spatial relations, (2) mapping coordinates to absolute directions, and (3) inferring user orientation. A curriculum learning strategy progressively builds these capabilities on Taiwan-LLM-13B-v2.0-Chat, a mid-sized model representative of resource-constrained settings. Experiments show that MCoT achieves 100% orientation accuracy on clean transcripts and 98.1% with ASR transcripts, substantially outperforming unimodal and non-structured baselines. Moreover, MCoT demonstrates robustness under noisy conversational conditions, including ASR recognition errors and multilingual code-switching. The model also maintains high accuracy in cross-domain evaluation and resilience to linguistic variation, domain shift, and referential ambiguity. These findings highlight the potential of structured MCoT spatial reasoning as a path toward interpretable and resource-efficient embodied navigation.
Authors: Yu Gu, Jingjing Fu, Xiaodong Liu, Jeya Maria Jose Valanarasu, Noel Codella, Reuben Tan, Qianchu Liu, Ying Jin, Sheng Zhang, Jinyu Wang, Rui Wang, Lei Song, Guanghui Qin, Naoto Usuyama, Cliff Wong, Cheng Hao, Hohin Lee, Praneeth Sanapathi, Sarah Hilado, Bian Jiang, Javier Alvarez-Valle, Mu Wei, Jianfeng Gao, Eric Horvitz, Matt Lungren, Hoifung Poon, Paul Vozila
Abstract: Large frontier models like GPT-5 now achieve top scores on medical benchmarks. But our stress tests tell a different story. Leading systems often guess correctly even when key inputs like images are removed, flip answers under trivial prompt changes, and fabricate convincing yet flawed reasoning. These aren't glitches; they expose how today's benchmarks reward test-taking tricks over medical understanding. We evaluate six flagship models across six widely used benchmarks and find that high leaderboard scores hide brittleness and shortcut learning. Through clinician-guided rubric evaluation, we show that benchmarks vary widely in what they truly measure yet are treated interchangeably, masking failure modes. We caution that medical benchmark scores do not directly reflect real-world readiness. If we want AI to earn trust in healthcare, we must demand more than leaderboard wins and must hold systems accountable for robustness, sound reasoning, and alignment with real medical demands.
Authors: Hongda Jiang, Xinyuan Zhang, Siddhant Garg, Rishab Arora, Shiun-Zu Kuo, Jiayang Xu, Christopher Brossman, Yue Liu, Aaron Colak, Ahmed Aly, Anuj Kumar, Xin Luna Dong
Abstract: We introduce Memory-QA, a novel real-world task that involves answering recall questions about visual content from previously stored multimodal memories. This task poses unique challenges, including the creation of task-oriented memories, the effective utilization of temporal and location information within memories, and the ability to draw upon multiple memories to answer a recall question. To address these challenges, we propose a comprehensive pipeline, Pensieve, integrating memory-specific augmentation, time- and location-aware multi-signal retrieval, and multi-memory QA fine-tuning. We created a multimodal benchmark to illustrate various real challenges in this task, and show the superior performance of Pensieve over state-of-the-art solutions (up to 14% on QA accuracy).
Authors: Seungyoun Shin, Dongha Ahn, Jiwoo Kim, Sungwook Jeon
Abstract: Recent work reports gains in neural text-to-speech (TTS) with Group Relative Policy Optimization (GRPO). However, in the absence of a verifiable reward for \textit{prosody}, GRPO trained on transcription-oriented signals (CER/NLL) lowers error rates yet collapses prosody into monotone, unnatural speech; adding speaker-similarity further destabilizes training and degrades CER. We address this with an \textit{iterative Direct Preference Optimization (DPO)} scheme that uses only a few hundred human-labeled preference pairs per round to directly optimize prosodic naturalness while regularizing to the current model. On \textbf{KoCC-TTS}, a curated dataset of authentic Korean call center interactions capturing task-oriented dialogues, our method attains the highest human preference (ELO) with competitive CER, outperforming GRPO and strong commercial baselines. These results suggest that when prosody cannot be rewarded automatically, \textit{human preference optimization} offers a practical and data-efficient path to natural and robust TTS. The demo page is available at \href{https://tts.ch.dev}
URLs: https://tts.ch.dev
Authors: Yuke Si, Runyan Yang, Yingying Gao, Junlan Feng, Chao Deng, Shilei Zhang
Abstract: Recent advances in large language models have facilitated the development of unified speech language models (SLMs) capable of supporting multiple speech tasks within a shared architecture. However, tasks such as automatic speech recognition (ASR) and speech emotion recognition (SER) rely on distinct types of information: ASR primarily depends on linguistic content, whereas SER requires the integration of both linguistic and paralinguistic cues. Existing multitask SLMs typically adopt naive parameter sharing or prompt-based conditioning without explicitly modeling the differences in information composition required by each task. Such designs risk task interference and performance degradation, especially under limited data conditions. To address these limitations, we propose HarmoniFuse, a component-selective and prompt-adaptive framework for multi-task speech language modeling. HarmoniFuse is designed to harmonize heterogeneous task demands by selecting and fusing task-relevant components of speech representations. Specifically, it integrates a gated speech encoder to extract task-specific acoustic features and a prompt-adaptive dynamic fusion module to aggregate transformer layers based on task characteristics. In addition, a batch-interleaved training strategy enables leveraging separate ASR and SER datasets without requiring joint annotation. Experimental results demonstrate that HarmoniFuse improves both ASR and SER performance, offering a scalable and robust solution for multitask speech understanding under realistic data constraints.
Authors: Runyan Yang, Yuke Si, Yingying Gao, Junlan Feng, Chao Deng, Shilei Zhang
Abstract: While large audio language models excel at tasks like ASR and emotion recognition, they still struggle with complex reasoning due to the modality gap between audio and text as well as the lack of structured intermediate supervision. To address this, we propose a unified knowledge distillation framework to transfer reasoning capabilities from a high-capacity textual teacher model to a student audio models while preserving its acoustic competence. Our method introduces two key dimensions: source-wise distillation, which leverages both textual and acoustic teachers to provide complementary modality-specific supervision; and layer-wise distillation, which aligns teacher signals with appropriate student layers to improve transfer efficiency. This dual-dimensional strategy enables fine-grained control over the distillation process, effectively bridging the gap between symbolic reasoning and speech representations. Experimental results show significant improvements in audio reasoning performance, demonstrating the effectiveness of our framework as a reasoning transfer solution for audio modeling.
Authors: Zhuoxiao Chen, Hongyang Yu, Ying Xu, Yadan Luo, Long Duong, Yuan-Fang Li
Abstract: Radiology report generation (RRG) aims to automatically produce clinically faithful reports from chest X-ray images. Prevailing work typically follows a scale-driven paradigm, by multi-stage training over large paired corpora and oversized backbones, making pipelines highly data- and compute-intensive. In this paper, we propose Oracle-educated GRPO {OraPO) with a FactScore-based reward (FactS) to tackle the RRG task under constrained budgets. OraPO enables single-stage, RL-only training by converting failed GRPO explorations on rare or difficult studies into direct preference supervision via a lightweight oracle step. FactS grounds learning in diagnostic evidence by extracting atomic clinical facts and checking entailment against ground-truth labels, yielding dense, interpretable sentence-level rewards. Together, OraPO and FactS create a compact and powerful framework that significantly improves learning efficiency on clinically challenging cases, setting the new SOTA performance on the CheXpert Plus dataset (0.341 in F1) with 2--3 orders of magnitude less training data using a small base VLM on modest hardware.
Authors: Yixin Liu, Yonghui Wu, Denghui Zhang, Lichao Sun
Abstract: The exponential growth of scientific literature poses unprecedented challenges for researchers attempting to synthesize knowledge across rapidly evolving fields. We present \textbf{Agentic AutoSurvey}, a multi-agent framework for automated survey generation that addresses fundamental limitations in existing approaches. Our system employs four specialized agents (Paper Search Specialist, Topic Mining \& Clustering, Academic Survey Writer, and Quality Evaluator) working in concert to generate comprehensive literature surveys with superior synthesis quality. Through experiments on six representative LLM research topics from COLM 2024 categories, we demonstrate that our multi-agent approach achieves significant improvements over existing baselines, scoring 8.18/10 compared to AutoSurvey's 4.77/10. The multi-agent architecture processes 75--443 papers per topic (847 total across six topics) while targeting high citation coverage (often $\geq$80\% on 75--100-paper sets; lower on very large sets such as RLHF) through specialized agent orchestration. Our 12-dimension evaluation captures organization, synthesis integration, and critical analysis beyond basic metrics. These findings demonstrate that multi-agent architectures represent a meaningful advancement for automated literature survey generation in rapidly evolving scientific domains.
Authors: Junyu Wang, Ziyang Ma, Zhengding Luo, Tianrui Wang, Meng Ge, Xiaobao Wang, Longbiao Wang
Abstract: Large Audio-Language Models (LALMs) often suffer from audio-textual attention imbalance, prioritizing text over acoustic information, particularly in the multi-modal fusion layers of the Transformer architecture. This bias hinders their ability to fully utilize acoustic cues, causing suboptimal performance on audio reasoning tasks. To mitigate this, we propose \textbf{MATA}, a novel training-free method that dynamically pushes LALMs to pay \textbf{M}ore \textbf{A}ttention \textbf{T}o \textbf{A}udio tokens within the self-attention mechanism. Specifically, MATA intervenes post raw attention scoring, targeting only the last token in intermediate layers without introducing additional parameters or computational overhead. Experiments on the MMAU and MMAR benchmarks confirm MATA's effectiveness, with consistent performance gains. Notably, on MMAR, MATA enables an open-source model to surpass the proprietary Gemini 2.0 Flash for the first time. Our work provides an efficient solution to mitigate attention bias and opens a new research direction for enhancing the audio-processing capabilities of multi-modal models.
Authors: Junhao Su, Yuanliang Wan, Junwei Yang, Hengyu Shi, Tianyang Han, Junfeng Luo, Yurui Qiu
Abstract: Tool-augmented large language models (LLMs) are usually trained with supervised imitation or coarse-grained reinforcement learning that optimizes single tool calls. Current self-reflection practices rely on heuristic prompts or one-way reasoning: the model is urged to 'think more' instead of learning error diagnosis and repair. This is fragile in multi-turn interactions; after a failure the model often repeats the same mistake. We propose structured reflection, which turns the path from error to repair into an explicit, controllable, and trainable action. The agent produces a short yet precise reflection: it diagnoses the failure using evidence from the previous step and then proposes a correct, executable follow-up call. For training we combine DAPO and GSPO objectives with a reward scheme tailored to tool use, optimizing the stepwise strategy Reflect, then Call, then Final. To evaluate, we introduce Tool-Reflection-Bench, a lightweight benchmark that programmatically checks structural validity, executability, parameter correctness, and result consistency. Tasks are built as mini trajectories of erroneous call, reflection, and corrected call, with disjoint train and test splits. Experiments on BFCL v3 and Tool-Reflection-Bench show large gains in multi-turn tool-call success and error recovery, and a reduction of redundant calls. These results indicate that making reflection explicit and optimizing it directly improves the reliability of tool interaction and offers a reproducible path for agents to learn from failure.
Authors: Hao Wang, Eiki Murata, Lingfang Zhang, Ayako Sato, So Fukuda, Ziqi Yin, Wentao Hu, Keisuke Nakao, Yusuke Nakamura, Sebastian Zwirner, Yi-Chia Chen, Hiroyuki Otomo, Hiroki Ouchi, Daisuke Kawahara
Abstract: Recent advances in multimodal large language models (MLLMs) have significantly enhanced video understanding capabilities, opening new possibilities for practical applications. Yet current video benchmarks focus largely on indoor scenes or short-range outdoor activities, leaving the challenges associated with long-distance travel largely unexplored. Mastering extended geospatial-temporal trajectories is critical for next-generation MLLMs, underpinning real-world tasks such as embodied-AI planning and navigation. To bridge this gap, we present VIR-Bench, a novel benchmark consisting of 200 travel videos that frames itinerary reconstruction as a challenging task designed to evaluate and push forward MLLMs' geospatial-temporal intelligence. Experimental results reveal that state-of-the-art MLLMs, including proprietary ones, struggle to achieve high scores, underscoring the difficulty of handling videos that span extended spatial and temporal scales. Moreover, we conduct an in-depth case study in which we develop a prototype travel-planning agent that leverages the insights gained from VIR-Bench. The agent's markedly improved itinerary recommendations verify that our evaluation protocol not only benchmarks models effectively but also translates into concrete performance gains in user-facing applications.
Authors: Zijian Ling, Han Zhang, Yazhuo Zhou, Jiahao Cui
Abstract: This paper presents ColorBlindnessEval, a novel benchmark designed to evaluate the robustness of Vision-Language Models (VLMs) in visually adversarial scenarios inspired by the Ishihara color blindness test. Our dataset comprises 500 Ishihara-like images featuring numbers from 0 to 99 with varying color combinations, challenging VLMs to accurately recognize numerical information embedded in complex visual patterns. We assess 9 VLMs using Yes/No and open-ended prompts and compare their performance with human participants. Our experiments reveal limitations in the models' ability to interpret numbers in adversarial contexts, highlighting prevalent hallucination issues. These findings underscore the need to improve the robustness of VLMs in complex visual environments. ColorBlindnessEval serves as a valuable tool for benchmarking and improving the reliability of VLMs in real-world applications where accuracy is critical.
Authors: Guoxin Wang, Jun Zhao, Xinyi Liu, Yanbo Liu, Xuyang Cao, Chao Li, Zhuoyun Liu, Qintian Sun, Fangru Zhou, Haoqiang Xing, Zhenhong Yang
Abstract: Medical imaging provides critical evidence for clinical diagnosis, treatment planning, and surgical decisions, yet most existing imaging models are narrowly focused and require multiple specialized networks, limiting their generalization. Although large-scale language and multimodal models exhibit strong reasoning and multi-task capabilities, real-world clinical applications demand precise visual grounding, multimodal integration, and chain-of-thought reasoning. We introduce Citrus-V, a multimodal medical foundation model that combines image analysis with textual reasoning. The model integrates detection, segmentation, and multimodal chain-of-thought reasoning, enabling pixel-level lesion localization, structured report generation, and physician-like diagnostic inference in a single framework. We propose a novel multimodal training approach and release a curated open-source data suite covering reasoning, detection, segmentation, and document understanding tasks. Evaluations demonstrate that Citrus-V outperforms existing open-source medical models and expert-level imaging systems across multiple benchmarks, delivering a unified pipeline from visual grounding to clinical reasoning and supporting precise lesion quantification, automated reporting, and reliable second opinions.
Authors: Karen Rosero, Eunjung Yeo, David R. Mortensen, Cortney Van't Slot, Rami R. Hallac, Carlos Busso
Abstract: We present ChiReSSD, a speech reconstruction framework that preserves children speaker's identity while suppressing mispronunciations. Unlike prior approaches trained on healthy adult speech, ChiReSSD adapts to the voices of children with speech sound disorders (SSD), with particular emphasis on pitch and prosody. We evaluate our method on the STAR dataset and report substantial improvements in lexical accuracy and speaker identity preservation. Furthermore, we automatically predict the phonetic content in the original and reconstructed pairs, where the proportion of corrected consonants is comparable to the percentage of correct consonants (PCC), a clinical speech assessment metric. Our experiments show Pearson correlation of 0.63 between automatic and human expert annotations, highlighting the potential to reduce the manual transcription burden. In addition, experiments on the TORGO dataset demonstrate effective generalization for reconstructing adult dysarthric speech. Our results indicate that disentangled, style-based TTS reconstruction can provide identity-preserving speech across diverse clinical populations.
Authors: Saeed Almheiri, Rania Hossam, Mena Attia, Chenxi Wang, Preslav Nakov, Timothy Baldwin, Fajri Koto
Abstract: Large language models (LLMs) often reflect Western-centric biases, limiting their effectiveness in diverse cultural contexts. Although some work has explored cultural alignment, the potential for cross-cultural transfer, using alignment in one culture to improve performance in others, remains underexplored. This paper investigates cross-cultural transfer of commonsense reasoning in the Arab world, where linguistic and historical similarities coexist with local cultural differences. Using a culturally grounded commonsense reasoning dataset covering 13 Arab countries, we evaluate lightweight alignment methods such as in-context learning and demonstration-based reinforcement (DITTO), alongside baselines like supervised fine-tuning and direct preference optimization. Our results show that merely 12 culture-specific examples from one country can improve performance in others by 10\% on average, within multilingual models. In addition, we demonstrate that out-of-culture demonstrations from Indonesia and US contexts can match or surpass in-culture alignment for MCQ reasoning, highlighting cultural commonsense transferability beyond the Arab world. These findings demonstrate that efficient cross-cultural alignment is possible and offer a promising approach to adapt LLMs to low-resource cultural settings.
Authors: Xidong Wang, Dingjie Song, Shunian Chen, Junyin Chen, Zhenyang Cai, Chen Zhang, Lichao Sun, Benyou Wang
Abstract: Expanding the long-context capabilities of Multi-modal Large Language Models~(MLLMs) is critical for advancing video understanding and high-resolution image analysis. Achieving this requires systematic improvements in model architecture, data construction, and training strategies, particularly to address challenges such as performance degradation with increasing image counts and high computational costs. In this paper, we propose a hybrid architecture that integrates Mamba and Transformer blocks, introduce data construction methods that capture both temporal and spatial dependencies, and employ a progressive training strategy. Our released model, LongLLaVA (\textbf{Long}-Context \textbf{L}arge \textbf{L}anguage \textbf{a}nd \textbf{V}ision \textbf{A}ssistant), demonstrates an effective balance between efficiency and performance. LongLLaVA achieves competitive results across various benchmarks while maintaining high throughput and low memory consumption. Notably, it can process nearly one thousand images on a single A100 80GB GPU, underscoring its potential for a wide range of multi-modal applications.
Authors: Ziyin Zhang, Hang Yu, Shijie Li, Peng Di, Jianguo Li, Rui Wang
Abstract: Programming languages possess rich semantic information - such as data flow - that is represented by graphs and not available from the surface form of source code. Recent code language models have scaled to billions of parameters, but model source code solely as text tokens while ignoring any other structural information. Conversely, models that do encode structural information of code make modifications to the Transformer architecture, limiting their scale and compatibility with pretrained LLMs. In this work, we take the best of both worlds with GALLa - Graph Aligned Large Language Models. GALLa utilizes graph neural networks and cross-modal alignment technologies to inject the structural information of code into LLMs as an auxiliary task during finetuning. This framework is both model-agnostic and task-agnostic, as it can be applied to any code LLM for any code downstream task, and requires the structural graph data only at training time from a corpus unrelated to the finetuning data, while incurring no cost at inference time over the baseline LLM. Experiments on five code tasks with seven different baseline LLMs ranging in size from 350M to 14B validate the effectiveness of GALLa, demonstrating consistent improvement over the baseline, even for powerful models such as LLaMA3 and Qwen2.5-Coder.
Authors: Peichao Lai, Zhengfeng Zhang, Wentao Zhang, Fangcheng Fu, Bin Cui
Abstract: Recently, using large language models (LLMs) for data augmentation has led to considerable improvements in unsupervised sentence embedding models. However, existing methods encounter two primary challenges: limited data diversity and high data noise. Current approaches often neglect fine-grained knowledge, such as entities and quantities, leading to insufficient diversity. Besides, unsupervised data frequently lacks discriminative information, and the generated synthetic samples may introduce noise. In this paper, we propose a pipeline-based data augmentation method via LLMs and introduce the Gaussian-decayed gradient-assisted Contrastive Sentence Embedding (GCSE) model to enhance unsupervised sentence embeddings. To tackle the issue of low data diversity, our pipeline utilizes knowledge graphs (KGs) to extract entities and quantities, enabling LLMs to generate more diverse samples. To address high data noise, the GCSE model uses a Gaussian-decayed function to limit the impact of false hard negative samples, enhancing the model's discriminative capability. Experimental results show that our approach achieves state-of-the-art performance in semantic textual similarity (STS) tasks, using fewer data samples and smaller LLMs, demonstrating its efficiency and robustness across various models.
Authors: Tunazzina Islam, Dan Goldwasser
Abstract: Climate change communication on social media increasingly employs microtargeting strategies to effectively reach and influence specific demographic groups. This study presents a post-hoc analysis of microtargeting practices within climate campaigns by leveraging large language models (LLMs) to examine Meta (previously known as Facebook) advertisements. Our analysis focuses on two key aspects: demographic targeting and fairness. We evaluate the ability of LLMs to accurately predict the intended demographic targets, such as gender and age group. Furthermore, we instruct the LLMs to generate explanations for their classifications, providing transparent reasoning behind each decision. These explanations reveal the specific thematic elements used to engage different demographic segments, highlighting distinct strategies tailored to various audiences. Our findings show that young adults are primarily targeted through messages emphasizing activism and environmental consciousness, while women are engaged through themes related to caregiving roles and social advocacy. Additionally, we conduct a comprehensive fairness analysis to uncover biases in model predictions. We assess disparities in accuracy and error rates across demographic groups using established fairness metrics such as Demographic Parity, Equal Opportunity, and Predictive Equality. Our findings indicate that while LLMs perform well overall, certain biases exist, particularly in the classification of male audiences. The analysis of thematic explanations uncovers recurring patterns in messaging strategies tailored to various demographic groups, while the fairness analysis underscores the need for more inclusive targeting methods. This study provides a valuable framework for future research aimed at enhancing transparency, accountability, and inclusivity in social media-driven climate campaigns.
Authors: Yedi Hu, Yunzhi Yao, Ningyu Zhang, Huajun Chen, Shumin Deng
Abstract: Model merging has emerged as a key technique for enhancing the capabilities and efficiency of Large Language Models (LLMs). The open-source community has driven model evolution by iteratively merging existing models, yet a principled understanding of the gains and underlying factors in model merging remains limited. In this work, we study model evolution through iterative merging, drawing an analogy to biological evolution, and introduce the concept of model kinship, the degree of similarity or relatedness between LLMs. Through comprehensive empirical analysis, we show that model kinship is closely linked to the performance improvements achieved by merging, providing a useful criterion for selecting candidate models. Building on this insight, we propose a new model merging strategy: Top-k Greedy Merging with Model Kinship, which can improve benchmark performance. Specifically, we discover that incorporating model kinship as a guiding criterion enables continuous merging while mitigating performance degradation caused by local optima, thereby facilitating more effective model evolution. Code is available at https://github.com/zjunlp/ModelKinship.
Authors: Lucius E. J. Bynum, Kyunghyun Cho
Abstract: In this work, we present sequence-driven structural causal models (SD-SCMs), a framework for specifying causal models with user-defined structure and language-model-defined mechanisms. We characterize how an SD-SCM enables sampling from observational, interventional, and counterfactual distributions according to the desired causal structure. We then leverage this procedure to propose a new type of benchmark for causal inference methods, generating individual-level counterfactual data to test treatment effect estimation. We create an example benchmark consisting of thousands of datasets, and test a suite of popular estimation methods for average, conditional average, and individual treatment effect estimation. We find under this benchmark that (1) causal methods outperform non-causal methods and that (2) even state-of-the-art methods struggle with individualized effect estimation, suggesting this benchmark captures some inherent difficulties in causal estimation. Apart from generating data, this same technique can underpin the auditing of language models for (un)desirable causal effects, such as misinformation or discrimination. We believe SD-SCMs can serve as a useful tool in any application that would benefit from sequential data with controllable causal structure.
Authors: Jisang Park, Minu Kim, DaYoung Hong, Jongha Lee
Abstract: Learners of a second language (L2) often map non-native phonemes with similar native-language (L1) phonemes, making conventional L2-focused training slow and effortful. To address this, we propose an L1-grounded pronunciation training method based on compositional phoneme approximation (CPA), a feature-based representation technique that approximates L2 sounds with sequences of L1 phonemes. Evaluations with 20 Korean non-native English speakers show that CPA-based training achieves a 76% in-box formant rate in acoustic analysis, over 20% relative improvement in phoneme recognition accuracy, and over 80% of speech being rated as more native-like, with minimal training.
Authors: Peichao Lai, Jiaxin Gan, Feiyang Ye, Yilei Wang, Bin Cui
Abstract: Sequence labeling remains a significant challenge in low-resource, domain-specific scenarios, particularly for character-dense languages like Chinese. Existing methods primarily focus on enhancing model comprehension and improving data diversity to boost performance. However, these approaches still struggle with inadequate model applicability and semantic distribution biases in domain-specific contexts. To overcome these limitations, we propose a novel framework that combines an LLM-based knowledge enhancement workflow with a span-based Knowledge Fusion for Rich and Efficient Extraction (KnowFREE) model. Our workflow employs explanation prompts to generate precise contextual interpretations of target entities, effectively mitigating semantic biases and enriching the model's contextual understanding. The KnowFREE model further integrates extension label features, enabling efficient nested entity extraction without relying on external knowledge during inference. Experiments on multiple Chinese domain-specific sequence labeling datasets demonstrate that our approach achieves state-of-the-art performance, effectively addressing the challenges posed by low-resource settings.
Authors: Shaina Raza, Ashmal Vayani, Aditya Jain, Aravind Narayanan, Vahid Reza Khazaie, Syed Raza Bashir, Elham Dolatabadi, Gias Uddin, Christos Emmanouilidis, Rizwan Qureshi, Mubarak Shah
Abstract: Detecting disinformation that blends manipulated text and images has become increasingly challenging, as AI tools make synthetic content easy to generate and disseminate. While most existing AI safety benchmarks focus on single modality misinformation (i.e., false content shared without intent to deceive), intentional multimodal disinformation, such as propaganda or conspiracy theories that imitate credible news, remains largely unaddressed. We introduce the Vision-Language Disinformation Detection Benchmark (VLDBench), the first large-scale resource supporting both unimodal (text-only) and multimodal (text + image) disinformation detection. VLDBench comprises approximately 62,000 labeled text-image pairs across 13 categories, curated from 58 news outlets. Using a semi-automated pipeline followed by expert review, 22 domain experts invested over 500 hours to produce high-quality annotations with substantial inter-annotator agreement. Evaluations of state-of-the-art Large Language Models (LLMs) and Vision-Language Models (VLMs) on VLDBench show that incorporating visual cues improves detection accuracy by 5 to 35 percentage points over text-only models. VLDBench provides data and code for evaluation, fine-tuning, and robustness testing to support disinformation analysis. Developed in alignment with AI governance frameworks (e.g., the MIT AI Risk Repository), VLDBench offers a principled foundation for advancing trustworthy disinformation detection in multimodal media. Project: https://vectorinstitute.github.io/VLDBench/ Dataset: https://huggingface.co/datasets/vector-institute/VLDBench Code: https://github.com/VectorInstitute/VLDBench
URLs: https://vectorinstitute.github.io/VLDBench/, https://huggingface.co/datasets/vector-institute/VLDBench, https://github.com/VectorInstitute/VLDBench
Authors: Dhananjay Ashok, Jonathan May
Abstract: The text produced by language models (LMs) can exhibit specific `behaviors,' such as a failure to follow alignment training, that we hope to detect and react to during deployment. Identifying these behaviors can often only be done post facto, i.e., after the entire text of the output has been generated. We provide evidence that there are times when we can predict how an LM will behave early in computation, before even a single token is generated. We show that probes trained on the internal representation of input tokens alone can predict a wide range of eventual behaviors over the entire output sequence. Using methods from conformal prediction, we provide provable bounds on the estimation error of our probes, creating precise early warning systems for these behaviors. The conformal probes can identify instances that will trigger alignment failures (jailbreaking) and instruction-following failures, without requiring a single token to be generated. An early warning system built on the probes reduces jailbreaking by 91%. Our probes also show promise in pre-emptively estimating how confident the model will be in its response, a behavior that cannot be detected using the output text alone. Conformal probes can preemptively estimate the final prediction of an LM that uses Chain-of-Thought (CoT) prompting, hence accelerating inference. When applied to an LM that uses CoT to perform text classification, the probes drastically reduce inference costs (65% on average across 27 datasets), with negligible accuracy loss. Encouragingly, probes generalize to unseen datasets and perform better on larger models, suggesting applicability to the largest of models in real-world settings.
Authors: Filippo Moment\`e, Alessandro Suglia, Mario Giulianelli, Ambra Ferrari, Alexander Koller, Oliver Lemon, David Schlangen, Raquel Fern\'andez, Raffaella Bernardi
Abstract: We examine three evaluation paradigms: standard benchmarks (e.g., MMLU and BBH), interactive games (e.g., Signalling Games or Taboo), and cognitive tests (e.g., for working memory or theory of mind). First, we investigate which of the former two-benchmarks or games-is most effective at discriminating LLMs of varying quality. Then, inspired by human cognitive assessments, we compile a suite of targeted tests that measure cognitive abilities deemed essential for effective language use, and we investigate their correlation with model performance in benchmarks and games. Our analyses reveal that interactive games are superior to standard benchmarks in discriminating models. Causal and logical reasoning correlate with both static and interactive tests, while differences emerge regarding core executive functions and social/emotional skills, which correlate more with games. We advocate for the development of new interactive benchmarks and targeted cognitive tasks inspired by assessing human abilities but designed specifically for LLMs.
Authors: Jintian Zhang, Yuqi Zhu, Mengshu Sun, Yujie Luo, Shuofei Qiao, Lun Du, Da Zheng, Huajun Chen, Ningyu Zhang
Abstract: Large language models (LLMs) have shown remarkable performance in complex reasoning tasks, but their efficiency is hindered by the substantial memory and computational costs associated with generating lengthy tokens. In this paper, we propose LightThinker, a novel method that enables LLMs to dynamically compress intermediate thoughts during reasoning. Inspired by human cognitive processes, LightThinker compresses verbose thought steps into compact representations and discards the original reasoning chains, thereby significantly reducing the number of tokens stored in the context window. This is achieved by training the model on when and how to perform compression through data construction, mapping hidden states to condensed gist tokens, and creating specialized attention masks. Additionally, we introduce the Dependency (Dep) metric to quantify the degree of compression by measuring the reliance on historical tokens during generation. Extensive experiments on four datasets and two models show that LightThinker reduces peak memory usage and inference time, while maintaining competitive accuracy. Our work provides a new direction for improving the efficiency of LLMs in complex reasoning tasks without sacrificing performance. Code is released at https://github.com/zjunlp/LightThinker.
Authors: Zahra Dehghanighobadi, Asja Fischer, Muhammad Bilal Zafar
Abstract: Explanations are an important tool for gaining insights into the behavior of ML models, calibrating user trust and ensuring regulatory compliance. Past few years have seen a flurry of post-hoc methods for generating model explanations, many of which involve computing model gradients or solving specially designed optimization problems. However, owing to the remarkable reasoning abilities of Large Language Model (LLMs), self-explanation, that is, prompting the model to explain its outputs has recently emerged as a new paradigm. In this work, we study a specific type of self-explanations, self-generated counterfactual explanations (SCEs). We design tests for measuring the efficacy of LLMs in generating SCEs. Analysis over various LLM families, model sizes, temperature settings, and datasets reveals that LLMs sometimes struggle to generate SCEs. Even when they do, their prediction often does not agree with their own counterfactual reasoning.
Authors: Xiulin Yang, Tatsuya Aoyama, Yuekun Yao, Ethan Wilcox
Abstract: Do language models (LMs) offer insights into human language learning? A common argument against this idea is that because their architecture and training paradigm are so vastly different from humans, LMs can learn arbitrary inputs as easily as natural languages. We test this claim by training LMs to model impossible and typologically unattested languages. Unlike previous work, which has focused exclusively on English, we conduct experiments on 12 languages from 4 language families with two newly constructed parallel corpora. Our results show that while GPT-2 small can largely distinguish attested languages from their impossible counterparts, it does not achieve perfect separation between all the attested languages and all the impossible ones. We further test whether GPT-2 small distinguishes typologically attested from unattested languages with different NP orders by manipulating word order based on Greenberg's Universal 20. We find that the model's perplexity scores do not distinguish attested vs. unattested word orders, while its performance on the generalization test does. These findings suggest that LMs exhibit some human-like inductive biases, though these biases are weaker than those found in human learners.
Authors: Tianyi Lorena Yan, Robin Jia
Abstract: To answer one-to-many factual queries (e.g., listing cities of a country), a language model (LM) must simultaneously recall knowledge and avoid repeating previous answers. How are these two subtasks implemented and integrated internally? Across multiple datasets, models, and prompt templates, we identify a promote-then-suppress mechanism: the model first recalls all answers, and then suppresses previously generated ones. Specifically, LMs use both the subject and previous answer tokens to perform knowledge recall, with attention propagating subject information and MLPs promoting the answers. Then, attention attends to and suppresses previous answer tokens, while MLPs amplify the suppression signal. Our mechanism is corroborated by extensive experimental evidence: in addition to using early decoding and causal tracing, we analyze how components use different tokens by introducing both Token Lens, which decodes aggregated attention updates from specified tokens, and a knockout method that analyzes changes in MLP outputs after removing attention to specified tokens. Overall, we provide new insights into how LMs' internal components interact with different input tokens to support complex factual recall. Code is available at https://github.com/Lorenayannnnn/how-lms-answer-one-to-many-factual-queries.
URLs: https://github.com/Lorenayannnnn/how-lms-answer-one-to-many-factual-queries.
Authors: Zhenyi Shen, Hanqi Yan, Linhai Zhang, Zhanghao Hu, Yali Du, Yulan He
Abstract: Chain-of-Thought (CoT) reasoning enhances Large Language Models (LLMs) by encouraging step-by-step reasoning in natural language. However, leveraging a latent continuous space for reasoning may offer benefits in terms of both efficiency and robustness. Prior implicit CoT methods attempt to bypass language completely by reasoning in continuous space but have consistently underperformed compared to the standard explicit CoT approach. We introduce CODI (Continuous Chain-of-Thought via Self-Distillation), a novel training framework that effectively compresses natural language CoT into continuous space. CODI jointly trains a teacher task (Explicit CoT) and a student task (Implicit CoT), distilling the reasoning ability from language into continuous space by aligning the hidden states of a designated token. Our experiments show that CODI is the first implicit CoT approach to match the performance of explicit CoT on GSM8k at the GPT-2 scale, achieving a 3.1x compression rate and outperforming the previous state-of-the-art by 28.2% in accuracy. CODI also demonstrates robustness, generalizable to complex datasets, and interpretability. These results validate that LLMs can reason effectively not only in natural language, but also in a latent continuous space. Code is available at https://github.com/zhenyi4/codi.
Authors: Yunzhi Yao, Jizhan Fang, Jia-Chen Gu, Ningyu Zhang, Shumin Deng, Huajun Chen, Nanyun Peng
Abstract: Knowledge Editing (KE) enables the modification of outdated or incorrect information in large language models (LLMs). While existing KE methods can update isolated facts, they often fail to generalize these updates to multi-hop reasoning tasks that rely on the modified knowledge. Through an analysis of reasoning circuits -- the neural pathways LLMs use for knowledge-based inference, we find that current layer-localized KE approaches (e.g., MEMIT, WISE), which edit only single or a few model layers, inadequately integrate updated knowledge into these reasoning pathways. To address this limitation, we present CaKE (Circuit-aware Knowledge Editing), a novel method that enhances the effective integration of updated knowledge in LLMs. By only leveraging a few curated data samples guided by our circuit-based analysis, CaKE stimulates the model to develop appropriate reasoning circuits for newly incorporated knowledge. Experiments show that CaKE enables more accurate and consistent use of edited knowledge across related reasoning tasks, achieving an average improvement of 20% in multi-hop reasoning accuracy on the MQuAKE dataset while requiring less memory than existing KE methods. We release the code and data in https://github.com/zjunlp/CaKE.
Authors: Kangwei Liu, Mengru Wang, Yujie Luo, Yuan Lin, Mengshu Sun, Lei Liang, Zhiqiang Zhang, Jun Zhou, Bryan Hooi, Shumin Deng
Abstract: Fine-tuning enables large language models (LLMs) to adapt to specific domains, but often compromises their previously established safety alignment. To mitigate the degradation of model safety during fine-tuning, we introduce LookAhead Tuning, a lightweight and effective data-driven approach that preserves safety during fine-tuning. The method introduces two simple strategies that modify training data by previewing partial answer prefixes, thereby minimizing perturbations to the model's initial token distributions and maintaining its built-in safety mechanisms. Comprehensive experiments demonstrate that LookAhead Tuning effectively maintains model safety without sacrificing robust performance on downstream tasks. Our findings position LookAhead Tuning as a reliable and efficient solution for the safe and effective adaptation of LLMs.
Authors: Yongrui Chen, Junhao He, Linbo Fu, Shenyu Zhang, Rihui Jin, Xinbang Dai, Jiaqi Li, Dehai Min, Nan Hu, Yuxin Zhang, Guilin Qi, Yi Huang, Tongtong Wu
Abstract: Unified Structured Knowledge Reasoning (USKR) aims to answer natural language questions (NLQs) by using structured sources such as tables, databases, and knowledge graphs in a unified way. Existing USKR methods either rely on employing task-specific strategies or custom-defined representations, which struggle to leverage the knowledge transfer between different SKR tasks or align with the prior of LLMs, thereby limiting their performance. This paper proposes a novel USKR framework named \textsc{Pandora}, which takes advantage of \textsc{Python}'s \textsc{Pandas} API to construct a unified knowledge representation for alignment with LLM pre-training. It employs an LLM to generate textual reasoning steps and executable Python code for each question. Demonstrations are drawn from a memory of training examples that cover various SKR tasks, facilitating knowledge transfer. Extensive experiments on four benchmarks involving three SKR tasks demonstrate that \textsc{Pandora} outperforms existing unified frameworks and competes effectively with task-specific methods.
Authors: Xinyue Lou, You Li, Jinan Xu, Xiangyu Shi, Chi Chen, Kaiyu Huang
Abstract: The rapid development of Multimodal Large Reasoning Models (MLRMs) has demonstrated broad application potential, yet their safety and reliability remain critical concerns that require systematic exploration. To address this gap, we conduct a comprehensive and systematic safety evaluation of 11 MLRMs across 5 benchmarks and unveil prevalent safety degradation phenomena in most advanced models. Moreover, our analysis reveals distinct safety patterns across different benchmarks: significant safety degradation is observed across jailbreak robustness benchmarks, whereas safety-awareness benchmarks demonstrate less pronounced degradation. In particular, the long thought process in some scenarios even enhances safety performance. Therefore, it is a potential approach to address safety issues in MLRMs by leveraging the intrinsic reasoning capabilities of the model to detect unsafe intent. To operationalize this insight, we construct a multimodal tuning dataset that incorporates a safety-oriented thought process. Experimental results from fine-tuning existing MLRMs with this dataset effectively enhances the safety on both jailbreak robustness and safety-awareness benchmarks. This study provides a new perspective for developing safe MLRMs. Our dataset is available at https://github.com/xinyuelou/Think-in-Safety.
Authors: Banca Calvo Figueras, Rodrigo Agerri
Abstract: The task of Critical Questions Generation (CQs-Gen) aims to foster critical thinking by enabling systems to generate questions that expose underlying assumptions and challenge the validity of argumentative reasoning structures. Despite growing interest in this area, progress has been hindered by the lack of suitable datasets and automatic evaluation standards. This paper presents a comprehensive approach to support the development and benchmarking of systems for this task. We construct the first large-scale dataset including ~5K manually annotated questions. We also investigate automatic evaluation methods and propose reference-based techniques as the strategy that best correlates with human judgments. Our zero-shot evaluation of 11 LLMs establishes a strong baseline while showcasing the difficulty of the task. Data and code plus a public leaderboard are provided to encourage further research, not only in terms of model performance, but also to explore the practical benefits of CQs-Gen for both automated reasoning and human critical thinking.
Authors: Md Mehrab Tanjim, Yeonjun In, Xiang Chen, Victor S. Bursztyn, Ryan A. Rossi, Sungchul Kim, Guang-Jie Ren, Vaishnavi Muppala, Shun Jiang, Yongsung Kim, Chanyoung Park
Abstract: Ambiguity remains a fundamental challenge in Natural Language Processing (NLP) due to the inherent complexity and flexibility of human language. With the advent of Large Language Models (LLMs), addressing ambiguity has become even more critical due to their expanded capabilities and applications. In the context of Conversational Question Answering (CQA), this paper explores the definition, forms, and implications of ambiguity for language driven systems, particularly in the context of LLMs. We define key terms and concepts, categorize various disambiguation approaches enabled by LLMs, and provide a comparative analysis of their advantages and disadvantages. We also explore publicly available datasets for benchmarking ambiguity detection and resolution techniques and highlight their relevance for ongoing research. Finally, we identify open problems and future research directions, especially in agentic settings, proposing areas for further investigation. By offering a comprehensive review of current research on ambiguities and disambiguation with LLMs, we aim to contribute to the development of more robust and reliable LLM-based systems.
Authors: Jinwang Song, Hongying Zan, Kunli Zhang, Lingling Mu, Yingjie Han, Haobo Hua, Min Peng
Abstract: Text-to-SQL, which maps natural language to SQL queries, has benefited greatly from recent advances in Large Language Models (LLMs). While LLMs offer various paradigms for this task, including prompting and supervised fine-tuning (SFT), SFT approaches still face challenges such as complex multi-stage pipelines and poor robustness to noisy schema information. To address these limitations, we present JOLT-SQL, a streamlined single-stage SFT framework that jointly optimizes schema linking and SQL generation via a unified loss. JOLT-SQL employs discriminative schema linking, enhanced by local bidirectional attention, alongside a confusion-aware noisy schema sampling strategy with selective attention to improve robustness under noisy schema conditions. Experiments on the Spider and BIRD benchmarks demonstrate that JOLT-SQL achieves state-of-the-art execution accuracy among comparable-size open-source models, while significantly improving both training and inference efficiency. Our code is available at https://github.com/Songjw133/JOLT-SQL.
Authors: DongGeon Lee, Joonwon Jang, Jihae Jeong, Hwanjo Yu
Abstract: Rapid deployment of vision-language models (VLMs) magnifies safety risks, yet most evaluations rely on artificial images. This study asks: How safe are current VLMs when confronted with meme images that ordinary users share? To investigate this question, we introduce MemeSafetyBench, a 50,430-instance benchmark pairing real meme images with both harmful and benign instructions. Using a comprehensive safety taxonomy and LLM-based instruction generation, we assess multiple VLMs across single and multi-turn interactions. We investigate how real-world memes influence harmful outputs, the mitigating effects of conversational context, and the relationship between model scale and safety metrics. Our findings demonstrate that VLMs are more vulnerable to meme-based harmful prompts than to synthetic or typographic images. Memes significantly increase harmful responses and decrease refusals compared to text-only inputs. Though multi-turn interactions provide partial mitigation, elevated vulnerability persists. These results highlight the need for ecologically valid evaluations and stronger safety mechanisms. MemeSafetyBench is publicly available at https://github.com/oneonlee/Meme-Safety-Bench.
Authors: Jisu Kim, Youngwoo Shin, Uiji Hwang, Jihun Choi, Richeng Xuan, Taeuk Kim
Abstract: Idioms have long posed a challenge due to their unique linguistic properties, which set them apart from other common expressions. While recent studies have leveraged large language models (LLMs) to handle idioms across various tasks, e.g., idiom-containing sentence generation and idiomatic machine translation, little is known about the underlying mechanisms of idiom processing in LLMs, particularly in multilingual settings. To this end, we introduce MIDAS, a new large-scale dataset of idioms in six languages, each paired with its corresponding meaning. Leveraging this resource, we conduct a comprehensive evaluation of LLMs' idiom processing ability, identifying key factors that influence their performance. Our findings suggest that LLMs rely not only on memorization, but also adopt a hybrid approach that integrates contextual cues and reasoning, especially when processing compositional idioms. This implies that idiom understanding in LLMs emerges from an interplay between internal knowledge retrieval and reasoning-based inference.
Authors: Prateek Verma, Mert Pilanci
Abstract: This paper presents a fascinating find: By training an auto-regressive LLM model on text tokens, the text model inherently develops internally an ability to understand images and audio, thereby developing the ability to see and hear just by reading. Popular audio and visual LLM models fine-tune text LLM models to give text output conditioned on images and audio embeddings. On the other hand, our architecture takes in patches of images, audio waveforms or tokens as input. It gives us the embeddings or category labels typical of a classification pipeline. We show the generality of text weights in aiding audio classification for datasets FSD-50K and GTZAN. Further, we show this working for image classification on CIFAR-10 and Fashion-MNIST, as well on image patches. This pushes the notion of text-LLMs learning powerful internal circuits that can be utilized by activating necessary connections for various applications rather than training models from scratch every single time.
Authors: Marcus Ma, Georgios Chochlakis, Niyantha Maruthu Pandiyan, Jesse Thomason, Shrikanth Narayanan
Abstract: Multi-label classification is prevalent in real-world settings, but the behavior of Large Language Models (LLMs) in this setting is understudied. We investigate how autoregressive LLMs perform multi-label classification, focusing on subjective tasks, by analyzing the output distributions of the models at each label generation step. We find that the initial probability distribution for the first label often does not reflect the eventual final output, even in terms of relative order and find LLMs tend to suppress all but one label at each generation step. We further observe that as model scale increases, their token distributions exhibit lower entropy and higher single-label confidence, but the internal relative ranking of the labels improves. Finetuning methods such as supervised finetuning and reinforcement learning amplify this phenomenon. We introduce the task of distribution alignment for multi-label settings: aligning LLM-derived label distributions with empirical distributions estimated from annotator responses in subjective tasks. We propose both zero-shot and supervised methods which improve both alignment and predictive performance over existing approaches. We find one method -- taking the max probability over all label generation distributions instead of just using the initial probability distribution -- improves both distribution alignment and overall F1 classification without adding any additional computation.
Authors: Abdellah El Mekki, Houdaifa Atou, Omer Nacar, Shady Shehata, Muhammad Abdul-Mageed
Abstract: Enhancing the linguistic capabilities of Large Language Models (LLMs) to include low-resource languages is a critical research area. Current research directions predominantly rely on synthetic data generated by translating English corpora, which, while demonstrating promising linguistic understanding and translation abilities, often results in models aligned with source language culture. These models frequently fail to represent the cultural heritage and values of local communities. This work proposes a methodology to create both synthetic and retrieval-based pre-training data tailored to a specific community, considering its (i) language, (ii) cultural heritage, and (iii) cultural values. We demonstrate our methodology using Egyptian and Moroccan dialects as testbeds, chosen for their linguistic and cultural richness and current underrepresentation in LLMs. As a proof-of-concept, we develop NileChat, a 3B parameter Egyptian and Moroccan Arabic LLM adapted for Egyptian and Moroccan communities, incorporating their language, cultural heritage, and values. Our results on various understanding, translation, and cultural and values alignment benchmarks show that NileChat outperforms existing Arabic-aware LLMs of similar size and performs on par with larger models. This work addresses Arabic dialect in LLMs with a focus on cultural and values alignment via controlled synthetic data generation and retrieval-augmented pre-training for Moroccan Darija and Egyptian Arabic, including Arabizi variants, advancing Arabic NLP for low-resource communities. We share our methods, data, and models with the community to promote the inclusion and coverage of more diverse communities in cultural LLM development: https://github.com/UBC-NLP/nilechat .
Authors: Yiyang Feng, Yichen Wang, Shaobo Cui, Boi Faltings, Mina Lee, Jiawei Zhou
Abstract: Large Language Models (LLMs) have demonstrated impressive capabilities in reasoning, positioning them as promising tools for supporting human problem-solving. However, what happens when their performance is affected by misinformation, i.e., incorrect inputs introduced by users due to oversights or gaps in knowledge? Such misinformation is prevalent in real-world interactions with LLMs, yet how it propagates within LLMs' reasoning process remains underexplored. Focusing on mathematical reasoning, we present a comprehensive analysis of how misinformation affects intermediate reasoning steps and final answers. We also examine how effectively LLMs can correct misinformation when explicitly instructed to do so. Even with explicit instructions, LLMs succeed less than half the time in rectifying misinformation, despite possessing correct internal knowledge, leading to significant accuracy drops (10.02% - 72.20%), and the degradation holds with thinking models (4.30% - 19.97%). Further analysis shows that applying factual corrections early in the reasoning process most effectively reduces misinformation propagation, and fine-tuning on synthesized data with early-stage corrections significantly improves reasoning factuality. Our work offers a practical approach to mitigating misinformation propagation.
Authors: Pingjun Hong, Beiduo Chen, Siyao Peng, Marie-Catherine de Marneffe, Barbara Plank
Abstract: There is increasing evidence of Human Label Variation (HLV) in Natural Language Inference (NLI), where annotators assign different labels to the same premise-hypothesis pair. However, within-label variation--cases where annotators agree on the same label but provide divergent reasoning--poses an additional and mostly overlooked challenge. Several NLI datasets contain highlighted words in the NLI item as explanations, but the same spans on the NLI item can be highlighted for different reasons, as evidenced by free-text explanations, which offer a window into annotators' reasoning. To systematically understand this problem and gain insight into the rationales behind NLI labels, we introduce LITEX, a linguistically-informed taxonomy for categorizing free-text explanations. Using this taxonomy, we annotate a subset of the e-SNLI dataset, validate the taxonomy's reliability, and analyze how it aligns with NLI labels, highlights, and explanations. We further assess the taxonomy's usefulness in explanation generation, demonstrating that conditioning generation on LITEX yields explanations that are linguistically closer to human explanations than those generated using only labels or highlights. Our approach thus not only captures within-label variation but also shows how taxonomy-guided generation for reasoning can bridge the gap between human and model explanations more effectively than existing strategies.
Authors: Di Wu, Seth Aycock, Christof Monz
Abstract: Large Language Models (LLMs) demonstrate strong reasoning capabilities for many tasks, often by explicitly decomposing the task via Chain-of-Thought (CoT) reasoning. Recent work on LLM-based translation designs hand-crafted prompts to decompose translation, or trains models to incorporate intermediate steps. Translating Step-by-step (Briakou et al., 2024), for instance, introduces a multi-step prompt with decomposition and refinement of translation with LLMs, which achieved state-of-the-art results on WMT24 test data. In this work, we scrutinise this strategy's effectiveness. Empirically, we find no clear evidence that performance gains stem from explicitly decomposing the translation process via CoT, at least for the models on test; and we show prompting LLMs to 'translate again' and self-refine yields even better results than human-like step-by-step prompting. While the decomposition influences translation behaviour, faithfulness to the decomposition has both positive and negative effects on translation. Our analysis therefore suggests a divergence between the optimal translation strategies for humans and LLMs.
Authors: Ho Yin 'Sam' Ng, Ting-Yao Hsu, Aashish Anantha Ramakrishnan, Branislav Kveton, Nedim Lipka, Franck Dernoncourt, Dongwon Lee, Tong Yu, Sungchul Kim, Ryan A. Rossi, Ting-Hao 'Kenneth' Huang
Abstract: Figure captions are crucial for helping readers understand and remember a figure's key message. Many models have been developed to generate these captions, helping authors compose better quality captions more easily. Yet, authors almost always need to revise generic AI-generated captions to match their writing style and the domain's style, highlighting the need for personalization. Despite language models' personalization (LaMP) advances, these technologies often focus on text-only settings and rarely address scenarios where both inputs and profiles are multimodal. This paper introduces LaMP-Cap, a dataset for personalized figure caption generation with multimodal figure profiles. For each target figure, LaMP-Cap provides not only the needed inputs, such as figure images, but also up to three other figures from the same document--each with its image, caption, and figure-mentioning paragraphs--as a profile to characterize the context. Experiments with four LLMs show that using profile information consistently helps generate captions closer to the original author-written ones. Ablation studies reveal that images in the profile are more helpful than figure-mentioning paragraphs, highlighting the advantage of using multimodal profiles over text-only ones.
Authors: Marija \v{S}akota, Robert West
Abstract: Many recent approaches to structured NLP tasks use an autoregressive language model $M$ to map unstructured input text $x$ to output text $y$ representing structured objects (such as tuples, lists, trees, code, etc.), where the desired output structure is enforced via constrained decoding. During training, these approaches do not require the model to be aware of the constraints, which are merely implicit in the training outputs $y$. This is advantageous as it allows for dynamic constraints without requiring retraining, but can lead to low-quality output during constrained decoding at test time. We overcome this problem with Boosted Constrained Decoding (BoostCD), which combines constrained and unconstrained decoding in two phases: Phase 1 decodes from the base model $M$ twice, in constrained and unconstrained mode, obtaining two weak predictions. In phase 2, a learned autoregressive boosted model combines the two weak predictions into one final prediction. The mistakes made by the base model with vs. without constraints tend to be complementary, which the boosted model learns to exploit for improved performance. We demonstrate the power of BoostCD by applying it to closed information extraction. Our model, BoostIE, outperforms prior approaches both in and out of distribution, addressing several common errors identified in those approaches.
Authors: Jonathan St-Onge, Ashley M. A. Fehr, Carter Ward, Calla G. Beauregard, Michael V. Arnold, Samuel F. Rosenblatt, Benjamin Cooley, Christopher M. Danforth, Peter Sheridan Dodds
Abstract: Describing and comparing complex systems requires principled, theoretically grounded tools. Built around the phenomenon of type turbulence, allotaxonographs provide map-and-list visual comparisons of pairs of heavy-tailed distributions. Allotaxonographs are designed to accommodate a wide range of instruments including rank- and probability-turbulence divergences, Jenson-Shannon divergence, and generalized entropy divergences. Here, we describe a suite of programmatic tools for rendering allotaxonographs for rank-turbulence divergence in Matlab, Javascript, and Python, all of which have different use cases.
Authors: Lyucheng Wu, Mengru Wang, Ziwen Xu, Tri Cao, Nay Oo, Bryan Hooi, Shumin Deng
Abstract: Recent progress in Multimodal Large Language Models (MLLMs) has unlocked powerful cross-modal reasoning abilities, but also raised new safety concerns, particularly when faced with adversarial multimodal inputs. To improve the safety of MLLMs during inference, we introduce a modular and adaptive inference-time intervention technology, AutoSteer, without requiring any fine-tuning of the underlying model. AutoSteer incorporates three core components: (1) a novel Safety Awareness Score (SAS) that automatically identifies the most safety-relevant distinctions among the model's internal layers; (2) an adaptive safety prober trained to estimate the likelihood of toxic outputs from intermediate representations; and (3) a lightweight Refusal Head that selectively intervenes to modulate generation when safety risks are detected. Experiments on LLaVA-OV and Chameleon across diverse safety-critical benchmarks demonstrate that AutoSteer significantly reduces the Attack Success Rate (ASR) for textual, visual, and cross-modal threats, while maintaining general abilities. These findings position AutoSteer as a practical, interpretable, and effective framework for safer deployment of multimodal AI systems.
Authors: Alva West, Luodan Zhang, Liuliu Zhang, Minjun Zhu, Yixuan Weng, Yue Zhang
Abstract: Large language models (LLMs) have shown the capability to generate fluent and logical content, presenting significant challenges to machine-generated text detection, particularly text polished by adversarial perturbations such as paraphrasing. Current zero-shot detectors often employ Gaussian distributions as statistical measure for computing detection thresholds, which falters when confronted with the heavy-tailed statistical artifacts characteristic of adversarial or non-native English texts. In this paper, we introduce T-Detect, a novel detection method that fundamentally redesigns the curvature-based detectors. Our primary innovation is the replacement of standard Gaussian normalization with a heavy-tailed discrepancy score derived from the Student's t-distribution. This approach is theoretically grounded in the empirical observation that adversarial texts exhibit significant leptokurtosis, rendering traditional statistical assumptions inadequate. T-Detect computes a detection score by normalizing the log-likelihood of a passage against the expected moments of a t-distribution, providing superior resilience to statistical outliers. We validate our approach on the challenging RAID benchmark for adversarial text and the comprehensive HART dataset. Experiments show that T-Detect provides a consistent performance uplift over strong baselines, improving AUROC by up to 3.9\% in targeted domains. When integrated into a two-dimensional detection framework (CT), our method achieves state-of-the-art performance, with an AUROC of 0.926 on the Books domain of RAID. Our contributions are a new, theoretically-justified statistical foundation for text detection, an ablation-validated method that demonstrates superior robustness, and a comprehensive analysis of its performance under adversarial conditions. Ours code are released at https://github.com/ResearAI/t-detect.
Authors: Alva West, Yixuan Weng, Minjun Zhu, Luodan Zhang, Zhen Lin, Guangsheng Bao, Yue Zhang
Abstract: The field of AI-generated text detection has evolved from supervised classification to zero-shot statistical analysis. However, current approaches share a fundamental limitation: they aggregate token-level measurements into scalar scores, discarding positional information about where anomalies occur. Our empirical analysis reveals that AI-generated text exhibits significant non-stationarity, statistical properties vary by 73.8\% more between text segments compared to human writing. This discovery explains why existing detectors fail against localized adversarial perturbations that exploit this overlooked characteristic. We introduce Temporal Discrepancy Tomography (TDT), a novel detection paradigm that preserves positional information by reformulating detection as a signal processing task. TDT treats token-level discrepancies as a time-series signal and applies Continuous Wavelet Transform to generate a two-dimensional time-scale representation, capturing both the location and linguistic scale of statistical anomalies. On the RAID benchmark, TDT achieves 0.855 AUROC (7.1\% improvement over the best baseline). More importantly, TDT demonstrates robust performance on adversarial tasks, with 14.1\% AUROC improvement on HART Level 2 paraphrasing attacks. Despite its sophisticated analysis, TDT maintains practical efficiency with only 13\% computational overhead. Our work establishes non-stationarity as a fundamental characteristic of AI-generated text and demonstrates that preserving temporal dynamics is essential for robust detection.
Authors: Ting Cai, Stephen Sheen, AnHai Doan
Abstract: Expanding the abbreviated column names of tables, such as "esal" to "employee salary", is critical for many downstream NLP tasks for tabular data, such as NL2SQL, table QA, and keyword search. This problem arises in enterprises, domain sciences, government agencies, and more. In this paper, we make three contributions that significantly advance the state of the art. First, we show that the synthetic public data used by prior work has major limitations, and we introduce four new datasets in enterprise/science domains, with real-world abbreviations. Second, we show that accuracy measures used by prior work seriously undercount correct expansions, and we propose new synonym-aware measures that capture accuracy much more accurately. Finally, we develop Columbo, a powerful LLM-based solution that exploits context, rules, chain-of-thought reasoning, and token-level analysis. Extensive experiments show that Columbo significantly outperforms NameGuess, the current most advanced solution, by 4-29%, over five datasets. Columbo has been used in production on EDI, a major data lake for environmental sciences.
Authors: Bolian Li, Yanran Wu, Xinyu Luo, Ruqi Zhang
Abstract: Aligning large language models (LLMs) with human preferences has become a critical step in their development. Recent research has increasingly focused on test-time alignment, where additional compute is allocated during inference to enhance LLM safety and reasoning capabilities. However, these test-time alignment techniques often incur substantial inference costs, limiting their practical application. We are inspired by the speculative sampling acceleration, which leverages a small draft model to efficiently predict future tokens, to address the efficiency bottleneck of test-time alignment. We introduce the reward-shifted speculative sampling (SSS) algorithm, in which the draft model is aligned with human preferences, while the target model remains unchanged. We theoretically demonstrate that the distributional shift between the aligned draft model and the unaligned target model can be exploited to recover the RLHF optimal solution without actually obtaining it, by modifying the acceptance criterion and bonus token distribution. Our algorithm achieves superior gold reward scores at a significantly reduced inference cost in test-time weak-to-strong alignment experiments, thereby validating both its effectiveness and efficiency.
Authors: Zijie Wang, Eduardo Blanco
Abstract: People often ask questions with false assumptions, a type of question that does not have regular answers. Answering such questions requires first identifying the false assumptions. Large Language Models (LLMs) often generate misleading answers to these questions because of hallucinations. In this paper, we focus on identifying and answering questions with false assumptions in several domains. We first investigate whether the problem reduces to fact verification. Then, we present an approach leveraging external evidence to mitigate hallucinations. Experiments with five LLMs demonstrate that (1) incorporating retrieved evidence is beneficial and (2) generating and validating atomic assumptions yields more improvements and provides an interpretable answer by pinpointing the false assumptions.
Authors: Rapha\"el Merx, Hanna Suominen, Trevor Cohn, Ekaterina Vylomova
Abstract: In machine translation (MT), health is a high-stakes domain characterised by widespread deployment and domain-specific vocabulary. However, there is a lack of MT evaluation datasets for low-resource languages in this domain. To address this gap, we introduce OpenWHO, a document-level parallel corpus of 2,978 documents and 26,824 sentences from the World Health Organization's e-learning platform. Sourced from expert-authored, professionally translated materials shielded from web-crawling, OpenWHO spans a diverse range of over 20 languages, of which nine are low-resource. Leveraging this new resource, we evaluate modern large language models (LLMs) against traditional MT models. Our findings reveal that LLMs consistently outperform traditional MT models, with Gemini 2.5 Flash achieving a +4.79 ChrF point improvement over NLLB-54B on our low-resource test set. Further, we investigate how LLM context utilisation affects accuracy, finding that the benefits of document-level translation are most pronounced in specialised domains like health. We release the OpenWHO corpus to encourage further research into low-resource MT in the health domain.
Authors: Yuemei Xu, Kexin Xu, Jian Zhou, Ling Hu, Lin Gui
Abstract: The current Large Language Models (LLMs) face significant challenges in improving their performance on low-resource languages and urgently need data-efficient methods without costly fine-tuning. From the perspective of language-bridge, we propose a simple yet effective method, namely BridgeX-ICL, to improve the zero-shot Cross-lingual In-Context Learning (X-ICL) for low-resource languages. Unlike existing works focusing on language-specific neurons, BridgeX-ICL explores whether sharing neurons can improve cross-lingual performance in LLMs. We construct neuron probe data from the ground-truth MUSE bilingual dictionaries, and define a subset of language overlap neurons accordingly to ensure full activation of these anchored neurons. Subsequently, we propose an HSIC-based metric to quantify LLMs' internal linguistic spectrum based on overlapping neurons, guiding optimal bridge selection. The experiments conducted on 4 cross-lingual tasks and 15 language pairs from 7 diverse families, covering both high-low and moderate-low pairs, validate the effectiveness of BridgeX-ICL and offer empirical insights into the underlying multilingual mechanisms of LLMs. The code is publicly available at https://github.com/xuyuemei/BridgeX-ICL.
Authors: Jie Zhang, Changzai Pan, Kaiwen Wei, Sishi Xiong, Yu Zhao, Xiangyu Li, Jiaxin Peng, Xiaoyan Gu, Jian Yang, Wenhan Chang, Zhenhe Wu, Jiang Zhong, Shuangyong Song, Yongxiang Li, Xuelong Li
Abstract: Extensive research has been conducted to explore the capabilities of large language models (LLMs) in table reasoning. However, the essential task of transforming tables information into reports remains a significant challenge for industrial applications. This task is plagued by two critical issues: 1) the complexity and diversity of tables lead to suboptimal reasoning outcomes; and 2) existing table benchmarks lack the capacity to adequately assess the practical application of this task. To fill this gap, we propose the table-to-report task and construct a bilingual benchmark named T2R-bench, where the key information flow from the tables to the reports for this task. The benchmark comprises 457 industrial tables, all derived from real-world scenarios and encompassing 19 industry domains as well as 4 types of industrial tables. Furthermore, we propose an evaluation criteria to fairly measure the quality of report generation. The experiments on 25 widely-used LLMs reveal that even state-of-the-art models like Deepseek-R1 only achieves performance with 62.71 overall score, indicating that LLMs still have room for improvement on T2R-bench.
Authors: Hao Zhang, Mengsi Lyu, Zhuo Chen, Xingrun Xing, Yulong Ao, Yonghua Lin
Abstract: Large Language Models (LLMs) demonstrate exceptional capabilities across various tasks, but their deployment is constrained by high computational and memory costs. Model pruning provides an effective means to alleviate these demands. However, existing methods often ignore the characteristics of prefill-decode (PD) disaggregation in practice. In this paper, we propose a novel pruning method for PD disaggregation inference, enabling more precise and efficient block and KV Cache pruning. Our approach constructs pruning and distillation sets to perform iterative block removal independently for the prefill and decode stages, obtaining better pruning solutions. Moreover, we introduce a token-aware cache pruning mechanism that retains all KV Cache in the prefill stage but selectively reuses entries for the first and last token sequences in selected layers during decode, reducing communication costs with minimal overhead. Extensive experiments demonstrate that our approach consistently achieves strong performance in both PD disaggregation and PD unified settings without disaggregation. Under the same (default) settings, our method achieves improved performance and faster inference, along with a 4.95$\times$ reduction in data transmission bandwidth consumption.
Authors: Ilham Wicaksono, Zekun Wu, Rahul Patel, Theo King, Adriano Koshiyama, Philip Treleaven
Abstract: As large language models transition to agentic systems, current safety evaluation frameworks face critical gaps in assessing deployment-specific risks. We introduce AgentSeer, an observability-based evaluation framework that decomposes agentic executions into granular action and component graphs, enabling systematic agentic-situational assessment. Through cross-model validation on GPT-OSS-20B and Gemini-2.0-flash using HarmBench single turn and iterative refinement attacks, we demonstrate fundamental differences between model-level and agentic-level vulnerability profiles. Model-level evaluation reveals baseline differences: GPT-OSS-20B (39.47% ASR) versus Gemini-2.0-flash (50.00% ASR), with both models showing susceptibility to social engineering while maintaining logic-based attack resistance. However, agentic-level assessment exposes agent-specific risks invisible to traditional evaluation. We discover "agentic-only" vulnerabilities that emerge exclusively in agentic contexts, with tool-calling showing 24-60% higher ASR across both models. Cross-model analysis reveals universal agentic patterns, agent transfer operations as highest-risk tools, semantic rather than syntactic vulnerability mechanisms, and context-dependent attack effectiveness, alongside model-specific security profiles in absolute ASR levels and optimal injection strategies. Direct attack transfer from model-level to agentic contexts shows degraded performance (GPT-OSS-20B: 57% human injection ASR; Gemini-2.0-flash: 28%), while context-aware iterative attacks successfully compromise objectives that failed at model-level, confirming systematic evaluation gaps. These findings establish the urgent need for agentic-situation evaluation paradigms, with AgentSeer providing the standardized methodology and empirical validation.
Authors: Haokun Li, Yazhou Zhang, Jizhi Ding, Qiuchi Li, Peng Zhang
Abstract: With the rapid advancement of Multimodal Large Language Models (MLLMs), they have demonstrated exceptional capabilities across a variety of vision-language tasks. However, current evaluation benchmarks predominantly focus on objective visual question answering or captioning, inadequately assessing the models' ability to understand complex and subjective human emotions. To bridge this gap, we introduce EmoBench-Reddit, a novel, hierarchical benchmark for multimodal emotion understanding. The dataset comprises 350 meticulously curated samples from the social media platform Reddit, each containing an image, associated user-provided text, and an emotion category (sad, humor, sarcasm, happy) confirmed by user flairs. We designed a hierarchical task framework that progresses from basic perception to advanced cognition, with each data point featuring six multiple-choice questions and one open-ended question of increasing difficulty. Perception tasks evaluate the model's ability to identify basic visual elements (e.g., colors, objects), while cognition tasks require scene reasoning, intent understanding, and deep empathy integrating textual context. We ensured annotation quality through a combination of AI assistance (Claude 4) and manual verification.We conducted a comprehensive evaluation of nine leading MLLMs, including GPT-5, Gemini-2.5-pro, and GPT-4o, on EmoBench-Reddit.
Authors: Heng Zhang, Chengzhi Zhang
Abstract: The automated generation of research workflows is essential for improving the reproducibility of research and accelerating the paradigm of "AI for Science". However, existing methods typically extract merely fragmented procedural components and thus fail to capture complete research workflows. To address this gap, we propose an end-to-end framework that generates comprehensive, structured research workflows by mining full-text academic papers. As a case study in the Natural Language Processing (NLP) domain, our paragraph-centric approach first employs Positive-Unlabeled (PU) Learning with SciBERT to identify workflow-descriptive paragraphs, achieving an F1-score of 0.9772. Subsequently, we utilize Flan-T5 with prompt learning to generate workflow phrases from these paragraphs, yielding ROUGE-1, ROUGE-2, and ROUGE-L scores of 0.4543, 0.2877, and 0.4427, respectively. These phrases are then systematically categorized into data preparation, data processing, and data analysis stages using ChatGPT with few-shot learning, achieving a classification precision of 0.958. By mapping categorized phrases to their document locations in the documents, we finally generate readable visual flowcharts of the entire research workflows. This approach facilitates the analysis of workflows derived from an NLP corpus and reveals key methodological shifts over the past two decades, including the increasing emphasis on data analysis and the transition from feature engineering to ablation studies. Our work offers a validated technical framework for automated workflow generation, along with a novel, process-oriented perspective for the empirical investigation of evolving scientific paradigms. Source code and data are available at: https://github.com/ZH-heng/research_workflow.
Authors: Yujia Hu, Ming Shan Hee, Preslav Nakov, Roy Ka-Wei Lee
Abstract: The advancement of Large Language Models (LLMs) has transformed natural language processing; however, their safety mechanisms remain under-explored in low-resource, multilingual settings. Here, we aim to bridge this gap. In particular, we introduce \textsf{SGToxicGuard}, a novel dataset and evaluation framework for benchmarking LLM safety in Singapore's diverse linguistic context, including Singlish, Chinese, Malay, and Tamil. SGToxicGuard adopts a red-teaming approach to systematically probe LLM vulnerabilities in three real-world scenarios: \textit{conversation}, \textit{question-answering}, and \textit{content composition}. We conduct extensive experiments with state-of-the-art multilingual LLMs, and the results uncover critical gaps in their safety guardrails. By offering actionable insights into cultural sensitivity and toxicity mitigation, we lay the foundation for safer and more inclusive AI systems in linguistically diverse environments.\footnote{Link to the dataset: https://github.com/Social-AI-Studio/SGToxicGuard.} \textcolor{red}{Disclaimer: This paper contains sensitive content that may be disturbing to some readers.}
Authors: Charlott Jakob, David Harbecke, Patrick Parschan, Pia Wenzel Neves, Vera Schmitt
Abstract: Large Language Models are increasingly used in applications requiring objective assessment, which could be compromised by political bias. Many studies found preferences for left-leaning positions in LLMs, but downstream effects on tasks like fact-checking remain underexplored. In this study, we systematically investigate political bias through exchanging words with euphemisms or dysphemisms in German claims. We construct minimal pairs of factually equivalent claims that differ in political connotation, to assess the consistency of LLMs in classifying them as true or false. We evaluate six LLMs and find that, more than political leaning, the presence of judgmental words significantly influences truthfulness assessment. While a few models show tendencies of political bias, this is not mitigated by explicitly calling for objectivism in prompts. Warning: This paper contains content that may be offensive or upsetting.
Authors: Tsz Ting Chung, Lemao Liu, Mo Yu, Dit-Yan Yeung
Abstract: Logic reasoning in natural language has been recognized as an important measure of human intelligence for Large Language Models (LLMs). Popular benchmarks may entangle multiple reasoning skills and thus provide unfaithful evaluations on the logic reasoning skill. Meanwhile, existing logic reasoning benchmarks are limited in language diversity and their distributions are deviated from the distribution of an ideal logic reasoning benchmark, which may lead to biased evaluation results. This paper thereby proposes a new classical logic benchmark DivLogicEval, consisting of natural sentences composed of diverse statements in a counterintuitive way. To ensure a more reliable evaluation, we also introduce a new evaluation metric that mitigates the influence of bias and randomness inherent in LLMs. Through experiments, we demonstrate the extent to which logical reasoning is required to answer the questions in DivLogicEval and compare the performance of different popular LLMs in conducting logical reasoning.
Authors: Jane Luo, Xin Zhang, Steven Liu, Jie Wu, Yiming Huang, Yangyu Huang, Chengyu Yin, Ying Xin, Jianfeng Liu, Yuefeng Zhan, Hao Sun, Qi Chen, Scarlett Li, Mao Yang
Abstract: Large language models excel at function- and file-level code generation, yet generating complete repositories from scratch remains a fundamental challenge. This process demands coherent and reliable planning across proposal- and implementation-level stages, while natural language, due to its ambiguity and verbosity, is ill-suited for faithfully representing complex software structures. To address this, we introduce the Repository Planning Graph (RPG), a persistent representation that unifies proposal- and implementation-level planning by encoding capabilities, file structures, data flows, and functions in one graph. RPG replaces ambiguous natural language with an explicit blueprint, enabling long-horizon planning and scalable repository generation. Building on RPG, we develop ZeroRepo, a graph-driven framework for repository generation from scratch. It operates in three stages: proposal-level planning and implementation-level refinement to construct the graph, followed by graph-guided code generation with test validation. To evaluate this setting, we construct RepoCraft, a benchmark of six real-world projects with 1,052 tasks. On RepoCraft, ZeroRepo generates repositories averaging 36K Code Lines, roughly 3.9$\times$ the strongest baseline (Claude Code) and about 64$\times$ other baselines. It attains 81.5% functional coverage and a 69.7% pass rate, exceeding Claude Code by 27.3 and 35.8 percentage points, respectively. Further analysis shows that RPG models complex dependencies, enables progressively more sophisticated planning through near-linear scaling, and enhances LLM understanding of repositories, thereby accelerating agent localization.
Authors: Wenjie Lin, Hange Liu, Xutao Mao, Yingying Zhuang, Jingwei Shi, Xudong Han, Tianyu Shi, Jinrui Yang
Abstract: We present ParlAI Vote, an interactive system for exploring European Parliament debates and votes, and for testing LLMs on vote prediction and bias analysis. This platform connects debate topics, speeches, and roll-call outcomes, and includes rich demographic data such as gender, age, country, and political group. Users can browse debates, inspect linked speeches, compare real voting outcomes with predictions from frontier LLMs, and view error breakdowns by demographic group. Visualizing the EuroParlVote benchmark and its core tasks of gender classification and vote prediction, ParlAI Vote highlights systematic performance bias in state-of-the-art LLMs. The system unifies data, models, and visual analytics in a single interface, lowering the barrier for reproducing findings, auditing behavior, and running counterfactual scenarios. It supports research, education, and public engagement with legislative decision-making, while making clear both the strengths and the limitations of current LLMs in political analysis.
Authors: Byeongho Yu, Changhun Lee, Jungyu Jin, Eunhyeok Park
Abstract: To mitigate the hallucination problem in large language models, DoLa exploits early exit logits from the same model as a contrastive prior. However, we found that these early exit logits tend to be flat, low in magnitude, and fail to reflect meaningful contrasts. To address this, we propose PruneCD, a novel contrastive decoding method that constructs the amateur model via layer pruning rather than early exit. This design leads to more informative and well-aligned logits, enabling more effective contrastive decoding. Through qualitative and quantitative analyses, we demonstrate that PruneCD consistently improves factuality with minimal inference overhead, offering a robust and practical approach to mitigating hallucinations in LLMs.
Authors: Xiaoqiang Kang, Shengen Wu, Zimu Wang, Yilin Liu, Xiaobo Jin, Kaizhu Huang, Wei Wang, Yutao Yue, Xiaowei Huang, Qiufeng Wang
Abstract: Existing table understanding methods face challenges due to complex table structures and intricate logical reasoning. While supervised finetuning (SFT) dominates existing research, reinforcement learning (RL), such as Group Relative Policy Optimization (GRPO), has shown promise but struggled with low initial policy accuracy and coarse rewards in tabular contexts. In this paper, we introduce Table-R1, a three-stage RL framework that enhances multimodal table understanding through: (1) Warm-up that prompts initial perception and reasoning capabilities, (2) Perception Alignment GRPO (PA-GRPO), which employs continuous Tree-Edit-Distance Similarity (TEDS) rewards for recognizing table structures and contents, and (3) Hint-Completion GRPO (HC-GRPO), which utilizes fine-grained rewards of residual steps based on the hint-guided question. Extensive experiments demonstrate that Table-R1 can boost the model's table reasoning performance obviously on both held-in and held-out datasets, outperforming SFT and GRPO largely. Notably, Qwen2-VL-7B with Table-R1 surpasses larger specific table understanding models (e.g., Table-LLaVA 13B), even achieving comparable performance to the closed-source model GPT-4o on held-in datasets, demonstrating the efficacy of each stage of Table-R1 in overcoming initialization bottlenecks and reward sparsity, thereby advancing robust multimodal table understanding.
Authors: Yongrui Chen, Yi Huang, Yunchang Liu, Shenyu Zhang, Junhao He, Tongtong Wu, Guilin Qi, Tianxing Wu
Abstract: Continual Structured Knowledge Reasoning (CSKR) focuses on training models to handle sequential tasks, where each task involves translating natural language questions into structured queries grounded in structured knowledge. Existing general continual learning approaches face significant challenges when applied to this task, including poor generalization to heterogeneous structured knowledge and inefficient reasoning due to parameter growth as tasks increase. To address these limitations, we propose a novel CSKR framework, \textsc{K-DeCore}, which operates with a fixed number of tunable parameters. Unlike prior methods, \textsc{K-DeCore} introduces a knowledge decoupling mechanism that disentangles the reasoning process into task-specific and task-agnostic stages, effectively bridging the gaps across diverse tasks. Building on this foundation, \textsc{K-DeCore} integrates a dual-perspective memory consolidation mechanism for distinct stages and introduces a structure-guided pseudo-data synthesis strategy to further enhance the model's generalization capabilities. Extensive experiments on four benchmark datasets demonstrate the superiority of \textsc{K-DeCore} over existing continual learning methods across multiple metrics, leveraging various backbone large language models.
Authors: Hyesung Jeon, Seojune Lee, Beomseok Kang, Yulhwa Kim, Jae-Joon Kim
Abstract: The demand for efficient deployment of large language models (LLMs) has driven interest in quantization, which reduces inference cost, and parameter-efficient fine-tuning (PEFT), which lowers training overhead. This motivated the development of quantization-aware PEFT to produce accurate yet efficient quantized models. In this setting, reducing quantization error prior to fine-tuning is crucial for achieving high model accuracy. However, existing methods that rely on low-rank adaptation suffer from limited representational capacity. Recent Fourier-related transform (FT)-based adapters offer greater representational power than low-rank adapters, but their direct integration into quantized models often results in ineffective error reduction and increased computational overhead. To overcome these limitations, we propose QWHA, a method that integrates FT-based adapters into quantized models by employing the Walsh-Hadamard Transform (WHT) as the transform kernel, together with a novel adapter initialization scheme incorporating adaptive parameter selection and value refinement. We demonstrate that QWHA effectively mitigates quantization errors while facilitating fine-tuning, and that its design substantially reduces computational cost. Experimental results show that QWHA consistently outperforms baselines in low-bit quantization accuracy and achieves significant training speedups over existing FT-based adapters. The code is available at https://github.com/vantaa89/qwha.
Authors: Kavin R V, Pawan Goyal
Abstract: Standard language models employ unique, monolithic embeddings for each token, potentially limiting their ability to capture the multifaceted nature of word meanings. We investigate whether tokens can be more effectively represented through a compositional structure that accumulates diverse semantic facets. To explore this, we propose Aggregate Semantic Grouping (ASG), a novel approach leveraging Product Quantization (PQ). We apply ASG to standard transformer architectures (mBERT, XLM-R, mT5) and evaluate this representational scheme across diverse tasks (NLI, NER, QA), as well as a biomedical domain-specific benchmark (BC5CDR) using BioBERT. Our findings demonstrate that representing tokens compositionally via ASG achieves extreme compression in embedding parameters (0.4--0.5\%) while maintaining $>$95\% task performance relative to the base model, even in generative tasks and extends to both cross lingual transfer and domain-specific settings. These results validate the principle that tokens can be effectively modeled as combinations of shared semantic building blocks. ASG offers a simple yet concrete method for achieving this, showcasing how compositional representations can capture linguistic richness while enabling compact yet semantically rich models.
Authors: Lekkala Sai Teja, Annepaka Yadagiri, Partha Pakray, Chukhu Chunka, Mangadoddi Srikar Vardhan
Abstract: Generation of Artificial Intelligence (AI) texts in important works has become a common practice that can be used to misuse and abuse AI at various levels. Traditional AI detectors often rely on document-level classification, which struggles to identify AI content in hybrid or slightly edited texts designed to avoid detection, leading to concerns about the model's efficiency, which makes it hard to distinguish between human-written and AI-generated texts. A sentence-level sequence labeling model proposed to detect transitions between human- and AI-generated text, leveraging nuanced linguistic signals overlooked by document-level classifiers. By this method, detecting and segmenting AI and human-written text within a single document at the token-level granularity is achieved. Our model combines the state-of-the-art pre-trained Transformer models, incorporating Neural Networks (NN) and Conditional Random Fields (CRFs). This approach extends the power of transformers to extract semantic and syntactic patterns, and the neural network component to capture enhanced sequence-level representations, thereby improving the boundary predictions by the CRF layer, which enhances sequence recognition and further identification of the partition between Human- and AI-generated texts. The evaluation is performed on two publicly available benchmark datasets containing collaborative human and AI-generated texts. Our experimental comparisons are with zero-shot detectors and the existing state-of-the-art models, along with rigorous ablation studies to justify that this approach, in particular, can accurately detect the spans of AI texts in a completely collaborative text. All our source code and the processed datasets are available in our GitHub repository.
Authors: Luigi Celona, Simone Bianco, Marco Donzella, Paolo Napoletano
Abstract: State-of-The-Art (SoTA) image captioning models are often trained on the MicroSoft Common Objects in Context (MS-COCO) dataset, which contains human-annotated captions with an average length of approximately ten tokens. Although effective for general scene understanding, these short captions often fail to capture complex scenes and convey detailed information. Moreover, captioning models tend to exhibit bias towards the ``average'' caption, which captures only the more general aspects, thus overlooking finer details. In this paper, we present a novel approach to generate richer and more informative image captions by combining the captions generated from different SoTA captioning models. Our proposed method requires no additional model training: given an image, it leverages pre-trained models from the literature to generate the initial captions, and then ranks them using a newly introduced image-text-based metric, which we name BLIPScore. Subsequently, the top two captions are fused using a Large Language Model (LLM) to produce the final, more detailed description. Experimental results on the MS-COCO and Flickr30k test sets demonstrate the effectiveness of our approach in terms of caption-image alignment and hallucination reduction according to the ALOHa, CAPTURE, and Polos metrics. A subjective study lends additional support to these results, suggesting that the captions produced by our model are generally perceived as more consistent with human judgment. By combining the strengths of diverse SoTA models, our method enhances the quality and appeal of image captions, bridging the gap between automated systems and the rich and informative nature of human-generated descriptions. This advance enables the generation of more suitable captions for the training of both vision-language and captioning models.
Authors: Brando Miranda, Patrick Yu, Saumya Goyal, Yu-Xiong Wang, Sanmi Koyejo
Abstract: In the context of few-shot learning, it is currently believed that a fixed pre-trained (PT) model, along with fine-tuning the final layer during evaluation, outperforms standard meta-learning algorithms. We re-evaluate these claims under an in-depth empirical examination of an extensive set of formally diverse datasets and compare PT to Model Agnostic Meta-Learning (MAML). Unlike previous work, we emphasize a fair comparison by using: the same architecture, the same optimizer, and all models trained to convergence. Crucially, we use a more rigorous statistical tool -- the effect size (Cohen's d) -- to determine the practical significance of the difference between a model trained with PT vs. a MAML. We then use a previously proposed metric -- the diversity coefficient -- to compute the average formal diversity of a dataset. Using this analysis, we demonstrate the following: 1. when the formal diversity of a data set is low, PT beats MAML on average and 2. when the formal diversity is high, MAML beats PT on average. The caveat is that the magnitude of the average difference between a PT vs. MAML using the effect size is low (according to classical statistical thresholds) -- less than 0.2. Nevertheless, this observation is contrary to the currently held belief that a pre-trained model is always better than a meta-learning model. Our extensive experiments consider 21 few-shot learning benchmarks, including the large-scale few-shot learning dataset Meta-Data set. We also show no significant difference between a MAML model vs. a PT model with GPT-2 on Openwebtext. We, therefore, conclude that a pre-trained model does not always beat a meta-learned model and that the formal diversity of a dataset is a driving factor.
Authors: Joseph Cho, Mrudang Mathur, Cyril Zakka, Dhamanpreet Kaur, Matthew Leipzig, Alex Dalal, Aravind Krishnan, Eubee Koo, Karen Wai, Cindy S. Zhao, Akshay Chaudhari, Matthew Duda, Ashley Choi, Ehsan Rahimy, Lyna Azzouz, Robyn Fong, Rohan Shad, William Hiesinger
Abstract: Deep learning algorithms require extensive data to achieve robust performance. However, data availability is often restricted in the medical domain due to patient privacy concerns. Synthetic data presents a possible solution to these challenges. Recently, image generative models have found increasing use for medical applications but are often designed for singular medical specialties and imaging modalities, thus limiting their broader utility. To address this, we introduce MediSyn: a text-guided, latent diffusion model capable of generating synthetic images from 6 medical specialties and 10 image types. Through extensive experimentation, we first demonstrate that MediSyn quantitatively matches or surpasses the performance of specialist models. Second, we show that our synthetic images are realistic and exhibit strong alignment with their corresponding text prompts, as validated by a team of expert physicians. Third, we provide empirical evidence that our synthetic images are visually distinct from their corresponding real patient images. Finally, we demonstrate that in data-limited settings, classifiers trained solely on synthetic data or real data supplemented with synthetic data can outperform those trained solely on real data. Our findings highlight the immense potential of generalist image generative models to accelerate algorithmic research and development in medicine.
Authors: Zongbo Han, Jialong Yang, Guangyu Wang, Junfan Li, Qianli Xu, Mike Zheng Shou, Changqing Zhang
Abstract: Vision-language foundation models (VLMs), such as CLIP, exhibit remarkable performance across a wide range of tasks. However, deploying these models can be unreliable when significant distribution gaps exist between training and test data, while fine-tuning for diverse scenarios is often costly. Cache-based test-time adapters offer an efficient alternative by storing representative test samples to guide subsequent classifications. Yet, these methods typically employ naive cache management with limited capacity, leading to severe catastrophic forgetting when samples are inevitably dropped during updates. In this paper, we propose DOTA (DistributiOnal Test-time Adaptation), a simple yet effective method addressing this limitation. Crucially, instead of merely memorizing individual test samples, DOTA continuously estimates the underlying distribution of the test data stream. Test-time posterior probabilities are then computed using these dynamically estimated distributions via Bayes' theorem for adaptation. This distribution-centric approach enables the model to continually learn and adapt to the deployment environment. Extensive experiments validate that DOTA significantly mitigates forgetting and achieves state-of-the-art performance compared to existing methods.
Authors: Jyh-Jing Hwang, Runsheng Xu, Hubert Lin, Wei-Chih Hung, Jingwei Ji, Kristy Choi, Di Huang, Tong He, Paul Covington, Benjamin Sapp, Yin Zhou, James Guo, Dragomir Anguelov, Mingxing Tan
Abstract: We introduce EMMA, an End-to-end Multimodal Model for Autonomous driving. Built upon a multi-modal large language model foundation like Gemini, EMMA directly maps raw camera sensor data into various driving-specific outputs, including planner trajectories, perception objects, and road graph elements. EMMA maximizes the utility of world knowledge from the pre-trained large language models, by representing all non-sensor inputs (e.g. navigation instructions and ego vehicle status) and outputs (e.g. trajectories and 3D locations) as natural language text. This approach allows EMMA to jointly process various driving tasks in a unified language space, and generate the outputs for each task using task-specific prompts. Empirically, we demonstrate EMMA's effectiveness by achieving state-of-the-art performance in motion planning on nuScenes as well as competitive results on the Waymo Open Motion Dataset (WOMD). EMMA also yields competitive results for camera-primary 3D object detection on the Waymo Open Dataset (WOD). We show that co-training EMMA with planner trajectories, object detection, and road graph tasks yields improvements across all three domains, highlighting EMMA's potential as a generalist model for autonomous driving applications. We hope that our results will inspire research to further evolve the state of the art in autonomous driving model architectures.
Authors: Dong Shu, Haiyan Zhao, Jingyu Hu, Weiru Liu, Ali Payani, Lu Cheng, Mengnan Du
Abstract: Large Vision-Language Models (LVLMs) have demonstrated remarkable capabilities in processing both visual and textual information. However, the critical challenge of alignment between visual and textual representations is not fully understood. This survey presents a comprehensive examination of alignment and misalignment in LVLMs through an explainability lens. We first examine the fundamentals of alignment, exploring its representational and behavioral aspects, training methodologies, and theoretical foundations. We then analyze misalignment phenomena across three semantic levels: object, attribute, and relational misalignment. Our investigation reveals that misalignment emerges from challenges at multiple levels: the data level, the model level, and the inference level. We provide a comprehensive review of existing mitigation strategies, categorizing them into parameter-frozen and parameter-tuning approaches. Finally, we outline promising future research directions, emphasizing the need for standardized evaluation protocols and in-depth explainability studies.
Authors: Yueyan Li, Wenhao Gao, Caixia Yuan, Xiaojie Wang
Abstract: The study of mechanistic interpretability aims to reverse-engineer a model to explain its behaviors. While recent studies have focused on the static mechanism of a certain behavior, the learning dynamics inside a model remain to be explored. In this work, we develop a fine-tuning method for analyzing the mechanism behind learning. Inspired by the concept of intrinsic dimension, we view a model as a computational graph with redundancy for a specific task, and treat the fine-tuning process as a search for and optimization of a subgraph within this graph. Based on this hypothesis, we propose circuit-tuning, an algorithm that iteratively builds the subgraph for a specific task and updates the relevant parameters in a heuristic way. We first validate our hypothesis through a carefully designed experiment and provide a detailed analysis of the learning dynamics during fine-tuning. Subsequently, we conduct experiments on more complex tasks, demonstrating that circuit-tuning could strike a balance between the performance on the target task and the general capabilities. Our work offers a new analytical method for the dynamics of fine-tuning, provides new findings on the mechanisms behind the training process, and inspires the design of superior algorithms for the training of neural networks.
Authors: Zhuoyuan Mao, Mengjie Zhao, Qiyu Wu, Hiromi Wakaki, Yuki Mitsufuji
Abstract: Recent advancements in music large language models (LLMs) have significantly improved music understanding tasks, which involve the model's ability to analyze and interpret various musical elements. These improvements primarily focused on integrating both music and text inputs. However, the potential of incorporating additional modalities such as images, videos and textual music features to enhance music understanding remains unexplored. To bridge this gap, we propose DeepResonance, a multimodal music understanding LLM fine-tuned via multi-way instruction tuning with multi-way aligned music, text, image, and video data. To this end, we construct Music4way-MI2T, Music4way-MV2T, and Music4way-Any2T, three 4-way training and evaluation datasets designed to enable DeepResonance to integrate both visual and textual music feature content. We also introduce multi-sampled ImageBind embeddings and a pre-LLM fusion Transformer to enhance modality fusion prior to input into text LLMs, tailoring for multi-way instruction tuning. Our model achieves state-of-the-art performances across six music understanding tasks, highlighting the benefits of the auxiliary modalities and the structural superiority of DeepResonance. We open-source the codes, models and datasets we constructed: github.com/sony/DeepResonance.
Authors: Yujiao Yang, Jing Lian, Linhui Li
Abstract: Mixture-of-Experts (MoE) enhances model performance while maintaining computational efficiency, making it well-suited for large-scale applications. Conventional mixture-of-experts (MoE) architectures suffer from suboptimal coordination dynamics, where isolated expert operations expose the model to overfitting risks. Moreover, they have not been effectively extended to attention blocks, which limits further efficiency improvements. To tackle these issues, we propose Union-of-Experts (UoE), which decomposes the transformer model into an equivalent group of experts and applies a hierarchical routing mechanism to allocate input subspaces to specialized experts. Our approach advances MoE design with four key innovations: (1) Constructing expert groups by partitioning non-MoE models into functionally equivalent specialists (2) Developing a hierarchical routing paradigm that integrates patch-wise data selection and expert selection strategies. (3) Extending the MoE design to attention blocks. (4) Proposing a hardware-optimized parallelization scheme that exploits batched matrix multiplications for efficient expert computation. The experiments demonstrate that our UoE model surpasses Full Attention, state-of-the-art MoEs and efficient transformers in several tasks across image and natural language domains. In language modeling tasks, UoE achieves an average reduction of 2.38 in perplexity compared to the best-performing MoE method with only 76% of its FLOPs. In the Long Range Arena benchmark, it demonstrates an average score at least 0.68% higher than all comparison models, with only 50% of the FLOPs of the best MoE method. In image classification, it yields an average accuracy improvement of 1.75% over the best model while maintaining comparable FLOPs. The source codes are available at https://github.com/YujiaoYang-work/UoE.
Authors: Dong Shu, Xuansheng Wu, Haiyan Zhao, Daking Rai, Ziyu Yao, Ninghao Liu, Mengnan Du
Abstract: Large Language Models (LLMs) have transformed natural language processing, yet their internal mechanisms remain largely opaque. Recently, mechanistic interpretability has attracted significant attention from the research community as a means to understand the inner workings of LLMs. Among various mechanistic interpretability approaches, Sparse Autoencoders (SAEs) have emerged as a promising method due to their ability to disentangle the complex, superimposed features within LLMs into more interpretable components. This paper presents a comprehensive survey of SAEs for interpreting and understanding the internal workings of LLMs. Our major contributions include: (1) exploring the technical framework of SAEs, covering basic architecture, design improvements, and effective training strategies; (2) examining different approaches to explaining SAE features, categorized into input-based and output-based explanation methods; (3) discussing evaluation methods for assessing SAE performance, covering both structural and functional metrics; and (4) investigating real-world applications of SAEs in understanding and manipulating LLM behaviors.
Authors: Mingyang Chen, Linzhuang Sun, Tianpeng Li, Haoze Sun, Yijie Zhou, Chenzheng Zhu, Haofen Wang, Jeff Z. Pan, Wen Zhang, Huajun Chen, Fan Yang, Zenan Zhou, Weipeng Chen
Abstract: Large Language Models (LLMs) have shown remarkable capabilities in reasoning, exemplified by the success of OpenAI-o1 and DeepSeek-R1. However, integrating reasoning with external search processes remains challenging, especially for complex multi-hop questions requiring multiple retrieval steps. We propose ReSearch, a novel framework that trains LLMs to Reason with Search via reinforcement learning without using any supervised data on reasoning steps. Our approach treats search operations as integral components of the reasoning chain, where when and how to perform searches is guided by text-based thinking, and search results subsequently influence further reasoning. We train ReSearch on Qwen2.5-7B(-Instruct) and Qwen2.5-32B(-Instruct) models and conduct extensive experiments. Despite being trained on only one dataset, our models demonstrate strong generalizability across various benchmarks. Analysis reveals that ReSearch naturally elicits advanced reasoning capabilities such as reflection and self-correction during the reinforcement learning process.
Authors: Han Wu, Jie Yin
Abstract: Few-shot relational learning on knowledge graph (KGs) aims to perform reasoning over relations with only a few training examples. While current methods have focused primarily on leveraging specific relational information, rich semantics inherent in KGs have been largely overlooked. To bridge this gap, we propose PromptMeta, a novel prompted meta-learning framework that seamlessly integrates meta-semantics with relational information for few-shot relational learning. PromptMeta introduces two core innovations: (1) a Meta-Semantic Prompt (MSP) pool that learns and consolidates high-level meta-semantics shared across tasks, enabling effective knowledge transfer and adaptation to newly emerging relations; and (2) a learnable fusion mechanism that dynamically combines meta-semantics with task-specific relational information tailored to different few-shot tasks. Both components are optimized jointly with model parameters within a meta-learning framework. Extensive experiments and analyses on two real-world KG benchmarks validate the effectiveness of PromptMeta in adapting to new relations with limited supervision.
Authors: Dong Shu, Xuansheng Wu, Haiyan Zhao, Mengnan Du, Ninghao Liu
Abstract: Sparse Autoencoders (SAEs) have recently emerged as powerful tools for interpreting and steering the internal representations of large language models (LLMs). However, conventional approaches to analyzing SAEs typically rely solely on input-side activations, without considering the causal influence between each latent feature and the model's output. This work is built on two key hypotheses: (1) activated latents do not contribute equally to the construction of the model's output, and (2) only latents with high causal influence are effective for model steering. To validate these hypotheses, we propose Gradient Sparse Autoencoder (GradSAE), a simple yet effective method that identifies the most influential latents by incorporating output-side gradient information.
Authors: Jingyan Shen, Jiarui Yao, Rui Yang, Yifan Sun, Feng Luo, Rui Pan, Tong Zhang, Han Zhao
Abstract: Reward modeling is a key step in building safe foundation models when applying reinforcement learning from human feedback (RLHF) to align Large Language Models (LLMs). However, reward modeling based on the Bradley-Terry (BT) model assumes a global reward function, failing to capture the inherently diverse and heterogeneous human preferences. Hence, such oversimplification limits LLMs from supporting personalization and pluralistic alignment. Theoretically, we show that when human preferences follow a mixture distribution of diverse subgroups, a single BT model has an irreducible error. While existing solutions, such as multi-objective learning with fine-grained annotations, help address this issue, they are costly and constrained by predefined attributes, failing to fully capture the richness of human values. In this work, we introduce MiCRo, a two-stage framework that enhances personalized preference learning by leveraging large-scale binary preference datasets without requiring explicit fine-grained annotations. In the first stage, MiCRo introduces context-aware mixture modeling approach to capture diverse human preferences. In the second stage, MiCRo integrates an online routing strategy that dynamically adapts mixture weights based on specific context to resolve ambiguity, allowing for efficient and scalable preference adaptation with minimal additional supervision. Experiments on multiple preference datasets demonstrate that MiCRo effectively captures diverse human preferences and significantly improves downstream personalization.
Authors: Shuai Wang, Zhenhua Liu, Jiaheng Wei, Xuanwu Yin, Dong Li, Emad Barsoum
Abstract: We present Athena-PRM, a multimodal process reward model (PRM) designed to evaluate the reward score for each step in solving complex reasoning problems. Developing high-performance PRMs typically demands significant time and financial investment, primarily due to the necessity for step-level annotations of reasoning steps. Conventional automated labeling methods, such as Monte Carlo estimation, often produce noisy labels and incur substantial computational costs. To efficiently generate high-quality process-labeled data, we propose leveraging prediction consistency between weak and strong completers as a criterion for identifying reliable process labels. Remarkably, Athena-PRM demonstrates outstanding effectiveness across various scenarios and benchmarks with just 5,000 samples. Furthermore, we also develop two effective strategies to improve the performance of PRMs: ORM initialization and up-sampling for negative data. We validate our approach in three specific scenarios: verification for test time scaling, direct evaluation of reasoning step correctness, and reward ranked fine-tuning. Our Athena-PRM consistently achieves superior performance across multiple benchmarks and scenarios. Notably, when using Qwen2.5-VL-7B as the policy model, Athena-PRM enhances performance by 10.2 points on WeMath and 7.1 points on MathVista for test time scaling. Furthermore, Athena-PRM sets the state-of-the-art (SoTA) results in VisualProcessBench and outperforms the previous SoTA by 3.9 F1-score, showcasing its robust capability to accurately assess the correctness of the reasoning step. Additionally, utilizing Athena-PRM as the reward model, we develop Athena-7B with reward ranked fine-tuning and outperforms baseline with a significant margin on five benchmarks.
Authors: Suhas BN, Andrew M. Sherrill, Jyoti Alaparthi, Dominik Mattioli, Rosa I. Arriaga, Chris W. Wiese, Saeed Abdullah
Abstract: Prolonged Exposure (PE) therapy is an effective treatment for post-traumatic stress disorder (PTSD), but evaluating therapist fidelity remains labor-intensive due to the need for manual review of session recordings. We present a method for the automatic temporal localization of key PE fidelity elements, identifying their start and stop times, directly from session audio and transcripts. Our approach fine-tunes a large pre-trained audio-language model, Qwen2-Audio, using Low-Rank Adaptation (LoRA) to process focused 30-second windows of audio-transcript input. Fidelity labels for three core protocol phases, therapist orientation (P1), imaginal exposure (P2), and post-imaginal processing (P3), are generated via LLM-based prompting and verified by trained raters. The model is trained to predict normalized boundary offsets using soft supervision guided by task-specific prompts. On a dataset of 308 real PE sessions, our best configuration (LoRA rank 8, 30s windows) achieves a mean absolute error (MAE) of 5.3s across tasks, within typical rater tolerance for timestamp review, enabling practical fidelity QC. We further analyze the effects of window size and LoRA rank, highlighting the importance of context granularity and model adaptation. This work introduces a privacy-preserving, scalable framework for fidelity tracking in PE therapy, with potential to support clinician training, supervision, and quality assurance.
Authors: Yu Wang, Shiwan Zhao, Zhihu Wang, Ming Fan, Xicheng Zhang, Yubo Zhang, Zhengfan Wang, Heyuan Huang, Ting Liu
Abstract: The integration of external knowledge through Retrieval-Augmented Generation (RAG) has become foundational in enhancing large language models (LLMs) for knowledge-intensive tasks. However, existing RAG paradigms often overlook the cognitive step of applying knowledge, leaving a gap between retrieved facts and task-specific reasoning. In this work, we introduce RAG+, a principled and modular extension that explicitly incorporates application-aware reasoning into the RAG pipeline. RAG+ constructs a dual corpus consisting of knowledge and aligned application examples, created either manually or automatically, and retrieves both jointly during inference. This design enables LLMs not only to access relevant information but also to apply it within structured, goal-oriented reasoning processes. Experiments across mathematical, legal, and medical domains, conducted on multiple models, demonstrate that RAG+ consistently outperforms standard RAG variants, achieving average improvements of 3-5%, and peak gains up to 13.5% in complex scenarios. By bridging retrieval with actionable application, RAG+ advances a more cognitively grounded framework for knowledge integration, representing a step toward more interpretable and capable LLMs.
Authors: Anand Gokhale, Vaibhav Srivastava, Francesco Bullo
Abstract: Large language models (LLMs) have shown promise in zero-shot and single step reasoning and decision making problems, but in long horizon sequential planning tasks, their errors compound, often leading to unreliable or inefficient behavior. We introduce LogicGuard, a modular actor-critic architecture in which an LLM actor is guided by a trajectory level LLM critic that communicates through Linear Temporal Logic (LTL). Our setup combines the reasoning strengths of language models with the guarantees of formal logic. The actor selects high-level actions from natural language observations, while the critic analyzes full trajectories and proposes new LTL constraints that shield the actor from future unsafe or inefficient behavior. LogicGuard supports both fixed safety rules and adaptive, learned constraints, and is model-agnostic: any LLM-based planner can serve as the actor, with LogicGuard acting as a logic-generating wrapper. We formalize planning as graph traversal under symbolic constraints, allowing LogicGuard to analyze failed or suboptimal trajectories and generate new temporal logic rules that improve future behavior. To demonstrate generality, we evaluate LogicGuard across two distinct settings: short-horizon general tasks and long-horizon specialist tasks. On the Behavior benchmark of 100 household tasks, LogicGuard increases task completion rates by 25% over a baseline InnerMonologue planner. On the Minecraft diamond-mining task, which is long-horizon and requires multiple interdependent subgoals, LogicGuard improves both efficiency and safety compared to SayCan and InnerMonologue. These results show that enabling LLMs to supervise each other through temporal logic yields more reliable, efficient and safe decision-making for both embodied agents.
Authors: Shane Waxler, Paul Blazek, Davis White, Daniel Sneider, Kevin Chung, Mani Nagarathnam, Patrick Williams, Hank Voeller, Karen Wong, Matthew Swanhorst, Sheng Zhang, Naoto Usuyama, Cliff Wong, Tristan Naumann, Hoifung Poon, Andrew Loza, Daniella Meeker, Seth Hain, Rahul Shah
Abstract: Realizing personalized medicine at scale calls for methods that distill insights from longitudinal patient journeys, which can be viewed as a sequence of medical events. Foundation models pretrained on large-scale medical event data represent a promising direction for scaling real-world evidence generation and generalizing to diverse downstream tasks. Using Epic Cosmos, a dataset with medical events from de-identified longitudinal health records for 16.3 billion encounters over 300 million unique patient records from 310 health systems, we introduce the Comet models, a family of decoder-only transformer models pretrained on 118 million patients representing 115 billion discrete medical events (151 billion tokens). We present the largest scaling-law study of medical event data, establishing a methodology for pretraining and revealing power-law scaling relationships for compute, tokens, and model size. Consequently, we pretrained a series of compute-optimal models with up to 1 billion parameters. Conditioned on a patient's real-world history, Comet autoregressively predicts the next medical event to simulate patient health timelines. We studied 78 real-world tasks, including diagnosis prediction, disease prognosis, and healthcare operations. Remarkably for a foundation model with generic pretraining and simulation-based inference, Comet generally outperformed or matched task-specific supervised models on these tasks, without requiring task-specific fine-tuning or few-shot examples. Comet's predictive power consistently improves as the model and pretraining scale. Our results show that Comet, a generative medical event foundation model, can effectively capture complex clinical dynamics, providing an extensible and generalizable framework to support clinical decision-making, streamline healthcare operations, and improve patient outcomes.
Authors: Jongyeop Hyun, Bumsoo Kim
Abstract: Recent advancements in Large Language Models (LLMs) have significantly improved reasoning capabilities, with in-context learning (ICL) emerging as a key technique for adaptation without retraining. While previous works have focused on leveraging correct examples, recent research highlights the importance of learning from errors to enhance performance. However, existing methods lack a structured framework for analyzing and mitigating errors, particularly in Multimodal Large Language Models (MLLMs), where integrating visual and textual inputs adds complexity. To address this issue, we propose REFINE: Retrieval-Enhanced Feedback via In-context Neural Error-book, a teacher-student framework that systematically structures errors and provides targeted feedback. REFINE introduces three systematic queries to construct structured feedback -- Feed-Target, Feed-Check, and Feed-Path -- to enhance multimodal reasoning by prioritizing relevant visual information, diagnosing critical failure points, and formulating corrective actions. Unlike prior approaches that rely on redundant retrievals, REFINE optimizes structured feedback retrieval, improving inference efficiency, token usage, and scalability. Our results demonstrate substantial speedup, reduced computational costs, and successful generalization, highlighting REFINE's potential for enhancing multimodal reasoning.
Authors: Terry Yue Zhuo, Dingmin Wang, Hantian Ding, Varun Kumar, Zijian Wang
Abstract: Large language models (LLMs) have demonstrated exceptional capabilities when trained within executable runtime environments, notably excelling at software engineering tasks through verified feedback loops. Yet, scalable and generalizable execution-grounded environments remain scarce, limiting progress in training more capable ML agents. We introduce CTF-Dojo, the first large-scale executable runtime tailored for training LLMs with verifiable feedback, featuring 658 fully functional Capture-The-Flag (CTF)-style challenges containerized in Docker with guaranteed reproducibility. To enable rapid scaling without manual intervention, we develop CTF-Forge, an automated pipeline that transforms publicly available artifacts into ready-to-use execution environments in minutes, eliminating weeks of expert configuration traditionally required. We trained LLM-based agents on just 486 high-quality, execution-verified trajectories from CTF-Dojo, achieving up to 11.6% absolute gains over strong baselines across three competitive benchmarks: InterCode-CTF, NYU CTF Bench, and Cybench. Our best-performing 32B model reaches 31.9% Pass@1, establishing a new open-weight state-of-the-art that rivals frontier models like DeepSeek-V3-0324 and Gemini-2.5-Flash. By framing CTF-style tasks as a benchmark for executable-agent learning, CTF-Dojo demonstrates that execution-grounded training signals are not only effective but pivotal in advancing high-performance ML agents without dependence on costly proprietary systems.
Authors: Shuo Shao, Yiming Li, Yu He, Hongwei Yao, Wenyuan Yang, Dacheng Tao, Zhan Qin
Abstract: The broad capabilities and substantial resources required to train Large Language Models (LLMs) make them valuable intellectual property, yet they remain vulnerable to copyright infringement, such as unauthorized use and model theft. LLM fingerprinting, a non-intrusive technique that extracts and compares the distinctive features from LLMs to identify infringements, offers a promising solution to copyright auditing. However, its reliability remains uncertain due to the prevalence of diverse model modifications and the lack of standardized evaluation. In this SoK, we present the first comprehensive study of LLM fingerprinting. We introduce a unified framework and formal taxonomy that categorizes existing methods into white-box and black-box approaches, providing a structured overview of the state of the art. We further propose LeaFBench, the first systematic benchmark for evaluating LLM fingerprinting under realistic deployment scenarios. Built upon mainstream foundation models and comprising 149 distinct model instances, LeaFBench integrates 13 representative post-development techniques, spanning both parameter-altering methods (e.g., fine-tuning, quantization) and parameter-independent mechanisms (e.g., system prompts, RAG). Extensive experiments on LeaFBench reveal the strengths and weaknesses of existing methods, thereby outlining future research directions and critical open problems in this emerging field. The code is available at https://github.com/shaoshuo-ss/LeaFBench.
Authors: Hanyi Mao, Quanjia Xiao, Lei Pang, Haixiao Liu
Abstract: We propose FSPO (Fair Sequence Policy Optimization), a sequence-level reinforcement learning method for LLMs that enforces length-fair clipping on the importance-sampling (IS) weight. We study RL methods with sequence-level IS and identify a mismatch when PPO/GRPO-style clipping is transplanted to sequences: a fixed clip range systematically reweights short vs.\ long responses, distorting the optimization direction. FSPO introduces a simple remedy: we clip the sequence log-IS ratio with a band that scales as $\sqrt{L}$. Theoretically, we formalize length fairness via a Length Reweighting Error (LRE) and prove that small LRE yields a cosine directional guarantee between the clipped and true updates. Empirically, FSPO flattens clip rates across length bins, stabilizes training, and outperforms all baselines across multiple evaluation datasets on Qwen3-8B-Base model.
Authors: Alva West, Yixuan Weng, Minjun Zhu, Zhen Lin, Zhiyuan Ning, Yue Zhang
Abstract: Failure attribution in multi-agent systems -- pinpointing the exact step where a decisive error occurs -- is a critical yet unsolved challenge. Current methods treat this as a pattern recognition task over long conversation logs, leading to critically low step-level accuracy (below 17\%), which renders them impractical for debugging complex systems. Their core weakness is a fundamental inability to perform robust counterfactual reasoning: to determine if correcting a single action would have actually averted the task failure. To bridge this \emph{counterfactual inference gap}, we introduce Abduct-Act-Predict (A2P) Scaffolding, a novel agent framework that transforms failure attribution from pattern recognition into a structured causal inference task. A2P explicitly guides a large language model through a formal three-step reasoning process within a single inference pass: (1) Abduction, to infer the hidden root causes behind an agent's actions; (2) Action, to define a minimal corrective intervention; and (3) Prediction, to simulate the subsequent trajectory and verify if the intervention resolves the failure. This structured approach leverages the holistic context of the entire conversation while imposing a rigorous causal logic on the model's analysis. Our extensive experiments on the Who\&When benchmark demonstrate its efficacy. On the Algorithm-Generated dataset, A2P achieves 47.46\% step-level accuracy, a 2.85$\times$ improvement over the 16.67\% of the baseline. On the more complex Hand-Crafted dataset, it achieves 29.31\% step accuracy, a 2.43$\times$ improvement over the baseline's 12.07\%. By reframing the problem through a causal lens, A2P Scaffolding provides a robust, verifiable, and significantly more accurate solution for automated failure attribution. Ours code are released at https://github.com/ResearAI/A2P.
Authors: Bishnu Bhusal, Manoj Acharya, Ramneet Kaur, Colin Samplawski, Anirban Roy, Adam D. Cobb, Rohit Chadha, Susmit Jha
Abstract: Large language models (LLMs) have significantly transformed natural language understanding and generation, but they raise privacy concerns due to potential exposure of sensitive information. Studies have highlighted the risk of information leakage, where adversaries can extract sensitive information embedded in the prompts. In this work, we introduce a novel private prediction framework for generating high-quality synthetic text with strong privacy guarantees. Our approach leverages the Differential Privacy (DP) framework to ensure worst-case theoretical bounds on information leakage without requiring any fine-tuning of the underlying models. The proposed method performs inference on private records and aggregates the resulting per-token output distributions. This enables the generation of longer and coherent synthetic text while maintaining privacy guarantees. Additionally, we propose a simple blending operation that combines private and public inference to further enhance utility. Empirical evaluations demonstrate that our approach outperforms previous state-of-the-art methods on in-context-learning (ICL) tasks, making it a promising direction for privacy-preserving text generation while maintaining high utility.
Authors: Jialiang Kang, Han Shu, Wenshuo Li, Yingjie Zhai, Xinghao Chen
Abstract: Speculative decoding is a widely adopted technique for accelerating inference in large language models (LLMs), yet its application to vision-language models (VLMs) remains underexplored, with existing methods achieving only modest speedups (<1.5x). This gap is increasingly significant as multimodal capabilities become central to large-scale models. We hypothesize that large VLMs can effectively filter redundant image information layer by layer without compromising textual comprehension, whereas smaller draft models struggle to do so. To address this, we introduce Vision-Aware Speculative Decoding (ViSpec), a novel framework tailored for VLMs. ViSpec employs a lightweight vision adaptor module to compress image tokens into a compact representation, which is seamlessly integrated into the draft model's attention mechanism while preserving original image positional information. Additionally, we extract a global feature vector for each input image and augment all subsequent text tokens with this feature to enhance multimodal coherence. To overcome the scarcity of multimodal datasets with long assistant responses, we curate a specialized training dataset by repurposing existing datasets and generating extended outputs using the target VLM with modified prompts. Our training strategy mitigates the risk of the draft model exploiting direct access to the target model's hidden states, which could otherwise lead to shortcut learning when training solely on target model outputs. Extensive experiments validate ViSpec, achieving, to our knowledge, the first substantial speedup in VLM speculative decoding. Code is available at https://github.com/KangJialiang/ViSpec.
Authors: Om Naphade, Saksham Bansal, Parikshit Pareek
Abstract: Hyper-parameter Tuning (HPT) is a necessary step in machine learning (ML) pipelines but becomes computationally expensive and opaque with larger models. Recently, Large Language Models (LLMs) have been explored for HPT, yet most rely on models exceeding 100 billion parameters. We propose an Expert Block Framework for HPT using Small LLMs. At its core is the Trajectory Context Summarizer (TCS), a deterministic block that transforms raw training trajectories into structured context, enabling small LLMs to analyze optimization progress with reliability comparable to larger models. Using two locally-run LLMs (phi4:reasoning14B and qwen2.5-coder:32B) and a 10-trial budget, our TCS-enabled HPT pipeline achieves average performance within ~0.9 percentage points of GPT-4 across six diverse tasks.
Authors: Kang-il Lee, Jahyun Koo, Seunghyun Yoon, Minbeom Kim, Hyukhun Koh, Dongryeol Lee, Kyomin Jung
Abstract: We introduce transductive program synthesis, a new formulation of the program synthesis task that explicitly leverages test inputs during synthesis. While prior approaches to program synthesis--whether based on natural language descriptions or input-output examples--typically aim to generalize from training examples, they often struggle with robustness, especially in real-world settings where training examples are limited and test inputs involve various edge cases. To address this, we propose a novel framework that improves robustness by treating synthesis as an active learning over a finite hypothesis class defined by programs' outputs. We use an LLM to predict outputs for selected test inputs and eliminate inconsistent hypotheses, where the inputs are chosen via a greedy maximin algorithm to minimize the number of LLM queries required. We evaluate our approach on four benchmarks: Playgol, MBPP+, 1D-ARC, and programmatic world modeling on MiniGrid. We demonstrate that our method significantly improves program synthesis in both accuracy and efficiency. We release our code at https://github.com/klee972/SYNTRA.